Every source, and how far it goes

Every entry here is required to record the strongest published argument against it. That rule is met 100% of the time. Some of the shortfall is legitimate, because a crystal structure has no counter-argument, but the rest is reading that has not been done yet.

1. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
   Q152 activity-decile stratification of the AFF1-only and recurrent-core promoter sets

   Iqbal Lab oncofusion knowledge graph, beat Q152 · this project 2026
   used by 'AFF1-only' is a property of mid-activity promoters: zero in the bottom three deciles, zero in the top decile, and 81% of the 332 in deciles 6-8, A partner-exclusive promoter set needs a label-permutation null, not an independence one, An independence null OVERSTATES partner-exclusive recurrence 16-fold here - it is not a lower bound, and the brief that dispatched this beat said it was, At gene-body scale the AFF1-exclusive recurrent class is 6 against an exact institution-stratified null of 0.86 - but the singleton class is 63 against 74, so recurrence carries the signal and exclusivity alone does not, BLNK (SLP-65) - the B-cell receptor adapter, in the Q152 AFF1-only core and a known lineage-infidelity marker in myeloid leukaemia, EPHA7 - ephrin type-A receptor 7, a published direct promoter target of the ALL1/AF4 (KMT2A::AFF1) fusion protein, EPHA7 is the one core member with a published KMT2A::AFF1 promoter ChIP - and this beat's own data put it second on separation while denying it to the AF9 samples, In the deposit paper's own target-gene table, the AFF1-exclusive class is 64 genes against a permutation null of 2.3, and it survives call-rate matching, institution stratification, B-lineage restriction and patient collapse, IRX1 sits in the 2-of-8 tier - two different patients, not the same one twice - and is not in the recurrent core, Nine of the seventeen core promoters have a non-AFF1 sample within 10% of the threshold, and one has a non-AFF1 sample STRONGER than its own weakest AFF1 call, p279 and p318 are one patient in two lineage states, and collapsing them removes exactly one gene - NOP58 - from the recurrent core, PLANNED: margin-ranked partner-specificity scan over all 6,049 occupied-anywhere promoters, with a depth control, POU2AF1 (OCA-B / OBF-1 / BOB-1) - the B-cell-restricted OCT coactivator, in the Q152 AFF1-only core, SPRED1 - the RAS/MAPK brake whose deletion marks relapse in paediatric B-ALL, third-widest margin in the Q152 AFF1-only core, TBX21 (T-bet) - the Th1/NK master regulator, the third lymphoid gene in the Q152 AFF1-only core and the one that makes the B-lineage reading fail, The 17-promoter AFF1-only recurrent core is not more recurrence than a label permutation of the same matrix gives, The 332-promoter AFF1-only set is BELOW the count a random eight-sample arm produces (null 425), The cannot-be-occupied floor never binds at gene-body scale in any of the 27 samples, so q147's gene-scale rule is a bare 2.0x gate - and the max-versus-95th-percentile difference between q147 and q152 is immaterial there, The lymphoid look of the core does not survive an activity-matched null inside the same 332, and the pre-registered power check says 17 genes cannot carry the statistic anyway, The per-library 'cannot be occupied' floor spans 0.00x to 4.01x an average promoter, and one core gene's AFF1-exclusivity rests entirely on it, The same occupancy rule passes 0.61% of gene bodies and 10.75% of promoters, so IRX1 at 3 of 27 is 18x its gene-scale expectation where at promoter scale it was below it, The separation MARGIN, not the recurrence count, is the statistic that identifies a partner-specific promoter, Thirteen of q152's seventeen AFF1-only core promoters are absent from the deposit paper's whole target-gene table, and the four that are present are AFF1-exclusive there too, Two of the nineteen non-AFF1 samples are menin-inhibitor treated, so the specificity arm is effectively seventeen - and removing them does not rescue the core, Two of the seventeen core promoter windows are shared with a neighbouring TSS - PBLD overlaps HNRNPH3 by 38.2% and CCP110 overlaps GDE1 by 7.8%, WT1 - Wilms tumour 1, leukaemia antigen and MRD marker, and the widest-margin promoter in the Q152 AFF1-only core
2. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
   curated-set-selectivity-enrichment: is this graph's curated gene set enriched for KMT2A-selectivity relative to the genome, and what is the base rate for the anti-selective half?

   · Oncofusion knowledge graph (internal analysis) 2026
   used by 'Four consecutive beats found 40-50% of their genes anti-selective' was the null expectation, not a pattern - and the real test shows the curated set IS enriched for selectivity, p = 0.0016
3. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
   Human Protein Atlas, consensus tissue RNA nTPM across 51 normal tissues (rna_tissue_consensus.tsv); HPA release v25.1, fetched 2026-09-23

   · Human Protein Atlas 2026
   used by 'Not expressed on normal haematopoietic cells' is a claim about blood, and it has been read for thirty years as a claim about the body, A bulk normal-tissue expression value includes the white blood cells living inside that tissue - which is why CD19 'failed' a normal-tissue safety filter at 31.2 nTPM in small intestine, A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, An NG2-directed therapy meets pericytes - and unlike B-cell aplasia there is no replacement therapy for a damaged vascular wall, LAMP5 was proposed as an immunotherapy target for MLL-rearranged leukaemia without a normal-tissue filter; two atlases independently put it above 290 in the basal ganglia, and it is not partner-specific either, LAMP5-directed therapy for KMT2A-rearranged leukaemia would meet the basal ganglia, in a disease treated in infants and routinely treated into the CNS, NG2/CSPG4 is the one antigen in the whole surfaceome that KMT2A::AFF1 itself drives - and it is a pericyte and vascular protein, at 124.5 nTPM in normal blood vessel and 246 TPM in tibial artery, REFUTED IN THE CYCLE IT WAS WRITTEN - the proposal that fusion-driven and normal-tissue-clean surface antigens are disjoint in KMT2A::AFF1 because of what the fusion's output programme is; an expression-matched permutation says it is the base rate, The screen this beat was dispatched to run 'because it has never been run' was published in Haematologica in 2022, used the same instrument, and is cited in this graph already - in an entry filed under innate immunity
4. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
   analyses/kmt2a-aff1-surface-target-screen/ - PREREGISTRATION.md, run.py and results.json, this repository. Pre-registered before any expression value was inspected.

   · Oncofusion Knowledge Graph 2026
   used by 'Not expressed on normal haematopoietic cells' is a claim about blood, and it has been read for thirty years as a claim about the body, A bulk normal-tissue expression value includes the white blood cells living inside that tissue - which is why CD19 'failed' a normal-tissue safety filter at 31.2 nTPM in small intestine, A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, Cell-surface capture proteomics on an isogenic human KMT2A::AFF1 model against its matched normal progenitor - the experiment an mRNA screen cannot replace, Fusion-driven surface antigens are NOT more exposed in normal tissue than expression-matched controls - ratio 0.91, permutation p = 0.79 - so the screen's empty intersection is the base rate and not a property of the fusion, LAMP5 was proposed as an immunotherapy target for MLL-rearranged leukaemia without a normal-tissue filter; two atlases independently put it above 290 in the basal ganglia, and it is not partner-specific either, NG2/CSPG4 is the one antigen in the whole surfaceome that KMT2A::AFF1 itself drives - and it is a pericyte and vascular protein, at 124.5 nTPM in normal blood vessel and 246 TPM in tibial artery, REFUTED IN THE CYCLE IT WAS WRITTEN - the proposal that fusion-driven and normal-tissue-clean surface antigens are disjoint in KMT2A::AFF1 because of what the fusion's output programme is; an expression-matched permutation says it is the base rate, S1PR4 - the third parenchymally clean surface protein in the screen, a leukocyte-restricted GPCR, and not fusion-specific either, SIT1 - one of only three surface proteins in the screen that is both expressed in KMT2A::AFF1 lines and absent from normal parenchyma, and it is clean because it is leukocyte-restricted, not because it is leukaemic
5. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 41634341
   DOT1L provides transcriptional memory through PRC1.1 antagonism.

   Neville D, Ferguson DT, Heikamp EB, Lai Z, Magor GW, Lam C, et al. · Nat Cell Biol 2026
   used by 'Transcriptional memory' means two different things in this literature, and both of them are used about MLL leukaemia, A DOT1L inhibitor, which does not touch the menin-KMT2A interface, removes 70% of the fusion's enrichment from its own strongest sites - two thirds of what a menin inhibitor removes - so fusion chromatin residency is partly downstream of H3K79 methylation rather than only of menin, A menin inhibitor does not only take something away - it causes PRC1.1 to put something on, and the drug needs the cell's own silencing machinery to work, A PCGF1-null cell - a published resistance genotype - loses H3K79me2 at the most-evicted sites exactly as a control cell does (0.6123 against 0.5860), so the new stratified readout scores it as fully drug-engaged, and stratification does NOT close the resistance blind spot it was hoped to close, A second disease finds the same DOT1L-to-non-canonical-PRC1 axis in the same year - and the mark moves the opposite way, A time-resolved BRG1/RING1B/H2AK119ub CUT&RUN at BENC under menin inhibition - the inv(16) experiment, run on the right disease, Asked for the first time whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor: all three pre-registered criteria fire, and the set they fire on is 52 low-amplitude bins reaching 6x the genome median, 16 of them with no gene within 20 kb and 1 of 52 on a CpG island where 66% of the fusion's real sites are, H2AK119ub - the mark every PRC1.1 claim in this graph rests on, named in fifteen entries and written in none until now, H3K79 methylation is what keeps Polycomb off - and that, not gene silencing, is what a DOT1L inhibitor actually undoes, Half the PRC1.1 intersection was free all along: Zhou's locus set is a publisher supplementary table, and 159 of this graph's gene entries are in it, Hypothesis: the menin inhibitor's effect is inactivation in place, not eviction - the fusion keeps most of its sites and loses its output, In inv(16), disrupting the fusion interface swaps BRG1 for RING1B at three MYC enhancers - the cleanest version anywhere of the experiment nobody has run on menin, In murine MLL-AF9, knocking out PCGF1 changes the menin inhibitor's effect on H3K79me2 by 0.0166 of a share ratio and the DOT1L inhibitor's by 0.0101 - both smaller than the 0.0584 two control cultures of the same line already differ by, so the mark's drug response runs outside PRC1.1, Instrument check before any conclusion: reading a browser-track .tdf with a hand-written binner and ranking the result blind returns the Hoxa cluster in five of the top twenty bins, plus Meis1, Jmjd1c, Six1, Arid2, Ncor1 and Tsc22d2 - the canonical MLL-AF9 programme, which is what the track claims to be, KDM2B, the PRC1.1 CpG-island recruiter, sits at the 99.7th percentile of its own distribution on the fusion's strongest sites in DMSO and does not gain share under either a menin inhibitor or a DOT1L inhibitor - it holds position while comparable bins drift down, Measure H3K27me3 after a menin inhibitor - the free pharmacodynamic nobody has read, and the one experiment that separates two mechanisms, Menin inhibition collapses the serine/one-carbon/ASNS programme in KMT2A::AFF1 cells - PHGDH -3.24, PSAT1 -2.40, ASNS -1.72, flat in a wild-type line - so it should re-sensitise KMT2A-rearranged ALL to asparaginase, a combination PubMed has zero records for, Menin inhibition is a reshuffle, not a subtraction: four proteins are documented to GAIN chromatin or persist under the drug, and none of them is the fusion, Menin inhibition recruits PRC1 to the BENC MYC enhancer cluster and evicts BRG1, exactly as CBFbeta-SMMHC inhibition does in inv(16), Menin inhibitors should leave half the H3K79 methylation standing - and if they do not, the menin-interface model is wrong somewhere, Nobody has convincingly found the H3K79 eraser - and that constraint sets the clock on every DOT1L inhibitor, PHGDH (O43175) - the largest menin-SELECTIVE fall in the whole serine panel, and the least essential gene in it, PRC1.1 is not both an effector and a resistance gene by coincidence: the two published readings are one activity measured from opposite ends, Prime, wash, divide, rechallenge - the one experiment that would turn 'DOT1L provides transcriptional memory' into a measurement, and it doubles as a dosing-schedule study, Re-bound at full text: still no publication puts KDM2B's rDNA repression together with a KMT2A fusion's nucleolar residence - 27 records, 10 of them congress abstract books, the rest reviews - but KDM2B ChIP-seq in MLL-AF9 cells is now deposited, so the rDNA re-map has a second and better dataset, Re-bound at full text: still nobody has asked whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor - but the data to ask it are now deposited, 158 ChIP-seq samples including spike-in-normalised FLAG-MLLAF9 under VTP50469, so the class moves from never-measured to measured-not-analysed, Re-bound at full text: the H3K79me2 decay half-life is still unmeasured - one paper in 12.4 million full texts contains the phrase 'half-life of H3K79me2', it is an unnumbered assertion citing a 2010 review, and the MLL::AF9 degron time-course that could have measured it uses the word zero times, The KMT2A-fusion causal chain: six links, one measured ordering, and two arms that are parallel rather than sequential, The premise that GSE260742 'carries a ChIP-Rx spike-in' does not survive reading the deposit: zero of 158 sample records mention a spike-in, the only ChIP-Rx experiment in the paper is an H2AK119ub panel that is not identifiable among the deposited samples, and 152 of 158 samples ship only an unnormalised igvtools .tdf, The serine arm replicates in a third KMT2A::AFF1 line from an independent laboratory - PHGDH -2.77, PSAT1 -2.17, GPT2 -2.34, ASNS -1.16 - but the ISR output genes do not, and it survives PCGF1 and BCOR knockout, The washout data cannot say whether the serine/ASNS collapse is durable - the washout control never recovers, ASNS is undetectable in the line, and the housekeeping drifts in opposite directions in the two matrices, Three unnamed things at once: the DOT1L-PRC1.1 deposition names the fusion, the cell types and both inhibitors with concentrations, Titrate PRC1.1 capacity against menin-inhibitor response, and measure H2AK119ub on the same timecourse as H3K79me2, What fills the space: RING1B and BCOR arrive at the fusion's sites under BOTH drugs, but the DOT1L inhibitor puts them everywhere on the fusion's territory while the menin inhibitor puts them only in the quartile where the fusion is most completely evicted - and at Meis1 RING1B rises 16-fold
6. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41578402
   Absence of Ledgf in mouse brain affects the Kmt2a/b and polycomb balance, synaptic transmission and motor function.

   Debusschere L, Bentea E, et al. · Acta Neuropathol Commun 2026
   used by "Dispensable for haematopoiesis" is a summary of a real phenotype: halved white counts, perturbed HSC and progenitor numbers, and failing serial replating, The result that runs the wrong way: in Psip1-null cells Hoxa and Hoxd expression goes UP, because Psip1 also recruits Bmi1 and Ctbp1
7. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
   PubMed esearch and efetch, run 2026-09-19; 60-record sample of ENL[tiab] classified by regular-expression screening of title plus abstract followed by inspection of the unmatched records

   · NCBI E-utilities 2026
   used by "ENL" returns 1,109 PubMed records of which about 30% are the protein: a measured alias-precision failure
8. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
   PubMed esearch and efetch 2026-09-19: (MLLT3[tiab] OR AF9[tiab] OR "AF-9"[tiab]) AND (phosphorylation OR phosphorylated OR kinase OR phosphosite)[tiab] = 106 records; first 40 titles inspected

   · NCBI E-utilities 2026
   used by "ENL" returns 1,109 PubMed records of which about 30% are the protein: a measured alias-precision failure
9. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
   Positive-control calibration of the Skucha 2018 singleton class, run 2026-09-24 by Q180. Script analyses/q180-is-af4-specificity-a-depth-artefact/positive_controls.py, output positive_controls.json; column map asserted against the sheet's header row.

   · This repository (script and JSON committed; deterministic, no network) 2026
   used by "Seen in MLL-AF4 and no other bait" excludes AFF1, AFF4 and ELL — the three most AFF1-specific proteins in the experiment — and keeps ACADM and GLUD1, A dimension can be saturated in the corpus and empty in the entry named for it, and the reporter that measures coverage cannot tell the difference, A ratio-of-ratios combines four independent tracks and a pair ratio combines two, so its noise floor is wider by a factor of root two in log space - and cycles 101 and 102 scored two ratio-of-ratios against a pair floor. Applying the correction to the measured 9.4% peak floor gives 13.5%: VTP's 1.3% stays comfortably inside, and SGC0946's 22.8% stays outside but its margin falls from '2-6x outside' to 1.7x, A sharper test than complex enrichment: is the claimed partner's OBLIGATE partner also there? — and its fourth failure is the one protein the source paper named as its own non-selectivity control, Across all seven MLL-fusion baits, how deep a bait's fixed top-300 list runs into non-significance predicts how many "specific" partners it appears to have — Spearman +0.821, exact p = 0.034, and MLL-AF4 is the extreme on both axes, Against a correctly measured same-condition floor, a menin inhibitor's effect on H3K4me3, H3K79me2 and H3K27me3 is a genome-wide redistribution - up in the decile below the peaks, down monotonically to the sparsest decile - while its effect on the FUSION stays confined to the top 0.1%. The drug displaces the complex at the peaks and moves the marks everywhere, AKAP95 IDR1 co-condenses with the KMT2A N-terminal fragment - the one AKAP95 contact that lands inside the region every KMT2A fusion retains, AKAP95 reaches this disease by two separate routes - through WRAD/DPY30 and through the retained KMT2A fragment - and only the second one survives the fusion, ARID4B / SAP180 (Q4LE39) - the Sin3A-complex subunit that shares SIN3A's HDAC1-binding region, is in the MLL-fusion interactome, and is the second of only two epigenetic vulnerabilities a CRISPR screen found in KMT2A-rearranged ALL, Auditing the graph against the PI's brief, not against its own ledger (run), Cycle 102 declared before it ran that an input pair 'bounds library and coverage variation but NOT ChIP efficiency' and is therefore a LOWER BOUND - the size of that underestimate is now measured in human KMT2A::AFF1 MV4;11 and it is sixteen-fold at the peaks: 0.6% for an input pair against 9.4% for a menin ChIP pair over the same bins, in the same cells, on the same day, Cycle 104's M1 - 'no scoreable mark moves at D1, so the menin inhibitor's chromatin effect is proportionally a peak phenomenon' - DOES NOT SURVIVE. The same-condition floor at D1 is 2.4%, not the 10% borrowed from the top stratum, and all three scoreable marks exceed it (H3K4me3 9.8%, H3K79me2 7.8%, H3K27me3 4.8%). Two of the three exceed even GSE260742's OWN stratum-matched input floor, which was on disk the whole time. Cycle 104's alternative M2 fires instead, Cycle 104's open question - 'why do two marks break the floor at D10 and not at D1?' - has a false premise. Against the measured same-condition floor of 2.3% at D10, THREE marks break it there (H3K79me2 16.2%, H3K4me3 12.6%, H3K27me3 6.3%), and the same three break it at D1 too. There is no D1-versus-D10 contrast to explain; there is a continuous bulk gradient that a single borrowed 10% figure had chopped in two, Every IntAct number in this project is drawn through a query that misses half the record - and what it misses is the taspase1 fragments, GATAD2A / p66-alpha (Q86YP4) - the subunit that bridges NuRD's two modules, the best-ranked NuRD gene in this project's KMT2A-selectivity scan, and present in the pan-fusion interactome under MLL-AF9 but NOT the index fusion, GATAD2B / p66-beta (Q8WXI9) - the NuRD scaffold with the broadest presence in the pan-MLL-fusion interactome, five of seven baits and still not the index fusion, and a haploinsufficiency syndrome that caps any therapeutic idea, GLUD1 is a hexameric mitochondrial matrix dehydrogenase recovered on 11 spectra with non-zero control signal from a nuclear pulldown — and this corpus has already declined one KMT2A partner whose sole source was a cryo-EM structure of GLUD1 — declined, H3C1 stays rejected, and now with the two numbers the registry row lacked: its accession carries TEN gene names, and removing every substrate and ribosome symbol moves interactome coverage 43.8% -> 46.1%, H4K20me1 - the transcription- and replication-linked H4 state, whose gene-body level tracks Pol II; it HAS been profiled genome-wide in human haematopoietic stem cells twice, and a query scoped to 'leukemia' cannot see either dataset, HDAC1 and HDAC2 belong to at least five different corepressor complexes, so an HDAC1/2 observation names no complex - and this graph carried both enzymes without the scaffolds that would tell them apart, In all 24 read-queue rows inspected, the candidate is the BAIT and this project's protein is the prey - nobody has ever tagged KMT2A, AFF4, menin or PSIP1 and found any of them; plus seven bounded rejections, In human KMT2A::AFF1 MV4;11 (monocytic, not B-lineage), the vehicle-versus-vehicle noise floor is 9.4% at the top 0.1% of menin/MLL1 sites and 0.1-2.6% in every other stratum - so cycle 102's borrowed 4-10% band is about right AT THE PEAKS and four to five times too generous everywhere else, and it was applied everywhere, IntAct curates SEVEN MLL fusion proteins as first-class interactors - ENL, AFF4, AF9, GAS7, ELL, AF6 and CALM - and not the index one, King verification of q181: the epigenetics field is NOT hollow (1 of 218 empty), the mark regex has three further blind spots including chromatin accessibility in 25 entries, and my own attempted correction of q181 was wrong, KMT2A::AFF1 has no measured fusion-specific interactome: the one dataset that appears to supply one is a rank-selected tail in the wrong lineage, and the question is open rather than answered badly, KMT2A::AFF1–LEF1 is one AP-MS row (p 0.0261, 10|8 spectra) in a Jurkat T-ALL host, where LEF1 is a resident lineage factor; it is the only one of the six AF4-specific candidates with non-zero negative-control signal, and no Wnt co-factor was co-recovered — declined, LEF1 — the Wnt/β-catenin DNA-binding factor of the T-lineage, named as an MLL-AF4-specific interactor in a T-ALL cell line and with essentially no B-lineage or KMT2A literature, Lineage-split DepMap: neither USP2 nor MAML2 is a dependency in KMT2A-rearranged lines, and USP22 is, MBD3 (O95983) - the methyl-CpG-binding subunit that does NOT bind methylated DNA, and a validated KMT2A::AFF1-selective CRISPR vulnerability that is absent from every column of the pan-fusion interactome, MBTD1 - the H4K20me1/me2 reader inside NuA4/TIP60, the quiescence brake on adult haematopoietic stem cells, and the 3' partner of ZMYND11::MBTD1 whose methyl-lysine cage the fusion carries and does not use, Measured: the cell-type, cell-state and epigenetics literature for KMT2A and MEN1 was NOT missing from this corpus - 7 of 16 load-bearing papers were already cited, one of them 32 times, and 0 of 16 were in gene-kmt2a, MLL fusions engage NuRD's remodelling arm without its deacetylase arm - and if so, every 'NuRD is repressive' inference in this graph is pointing the wrong way, MTA2 / MTA1L1 (O94776) - the NuRD scaffold that clamps HDAC1, recruited in T lymphocytes by BCL11B, and absent together with MTA1 and MTA3 from every column of the pan-MLL-fusion interactome, No experiment has ever ASKED whether MTA2, MBD3, GATAD2A or SIN3A binds a KMT2A fusion - four of them were nonetheless in the lysate of a seven-fusion pulldown, and the class of that absence is measured-not-reported, not never-measured, No orthogonal validation exists for any of the 110 MLL-AF4-specific partners, as far as four declared bounds can see — and the one partner the paper itself followed up, SETD2, is on the opposite list, shared by all seven fusions, Only 2 of this graph's 26 bounded absences name a deposited dataset — the corpus asks whether a paper exists and almost never whether the data does, Only 69 of the 529 entries carrying absence_class are absence entries, and every conceptually novel token authors reached for - partially-measured, measured-and-contradicted, not-worth-measuring - is on an interface entry, where the thing being graded is not an absence at all, PLANNED: re-search the public PXD009338 raw files for the twelve corepressor subunits the published table truncated away - one compute job that closes twelve measured-not-reported absences and adjudicates the module-split hypothesis, Prediction 5 of hyp-q117 contains three separable claims - normal eviction, no silencing, and return on withdrawal - and the deposit it names supports one of them only through a clause it never states; the eviction and silencing clauses need tracks the NTvKO arm does not carry, and withdrawal is absent from all 158 samples, Reading the fusion interactome by COMPLEX rather than by protein — and the strongest signature in it is a known contaminant, which is what makes the rest readable, Repeat the one published RPA-at-the-KMT2A-breakpoint experiment in a haematopoietic cell, with an RPA-dose arm - the protocol is fully specified in a 2010 paper whose every arm used fibroblasts and a CML line, and whose primer sets, dose and controls are all published, RPA1 and RPA2 are curated HGNC aliases of POLR1A and POLR1B, and RPA34 - a UniProt synonym of RPA2 - is an alias of POLR1G, so three of the ssDNA clamp's names point at RNA polymerase I subunits; the hazard is in symbol-based DATABASE lookups, not in PubMed, and the measured contamination says which, RPA2 is the only compartment-compatible candidate of the six and the corpus already has a KMT2A resection thread it would fit — but it is an abundant ssDNA-binding subunit on 6 spectra in a benzonase-treated lysate with no RPA1 or RPA3 co-recovered — declined, thread flagged, scripts/is-symbol-covered.py returns 'no match' for BRAF, AKT1, EGFR, CD226 and IRF2 - all five have a declined interface entry. The `interface-` prefix in its PREFIXES list can never match anything., SIN3A (Q96ST3) - the scaffold of the OTHER HDAC1/2 corepressor complex, a 5' fusion partner in infant B-ALL and in myelodysplasia, and present in three MLL-fusion AP-MS columns but not the index fusion's, The 297 absence-bearing entries that are not named absence-* split cleanly by entry type: the 109 gene and protein entries are bounded by title/abstract search (7 of 7 hand-read), the analyses are bounded by deposited data (0 of 6), and the weighted abstract-level share is 60% - indistinguishable from the 59% measured on the 39 named ones, The corpus holds 42 of 42 pan-fusion partners and 1 of 110 MLL-AF4-specific ones — the inverse of the standing priority, and every mechanism that produced it was a reasonable local decision, The coverage checker the standing brief mandates overwrites its own HGNC index mid-loop, so every symbol after the first no-match is answered wrongly - and the wrong answer is 'treat any answer as unsafe', The interactome coverage figure is computed on 651 of the 1,015 IntAct rows available - 64% - and the gap is worst for menin and LEDGF, the two baits the menin-interface hypothesis rests on, The missing reagent is a spike-in, not a better statistic: a spike-in-normalised H3K4me3, H3K79me2 and H3K27me3 ChIP-seq at 72 h of a clinical menin inhibitor in a KMT2A-rearranged human line, which is the only design that can tell a genuine genome-wide mark redistribution from the compositional shadow of a peak-localised change, The MLL-fusion interactome recovers corepressor SCAFFOLDS and not their HDAC1/2 catalytic core: NuRD splits 5/7 remodelling arm versus 0/7 deacetylase arm (Fisher p = 0.021), and HDAC1, HDAC2, RBBP4 and RBBP7 are absent from all seven baits, The NuRD complex - two separable modules, three paralogue choices that change what it does, and a graph that carried its ATPase and its shared core but never the scaffolds that make it NuRD, The reopening watchlist read for the first time: 0 of 10 candidate papers meet their condition, 8 of 10 watched decisions returned nothing, and not one of the ten titles names the gene it was retrieved for, The rule-64 complex-form query that the RPA2 decline flagged as unrun has now been run, and it finds a published damage-state RPA occupancy measurement AT the KMT2A breakpoint cluster region - so two of this corpus's absence entries carry ids that are broader than the claims inside them, The selection control, run at last: two biologically identical vehicle-treated input libraries produce a MONOTONE decile gradient from 0.9040 to 1.2349 - wider than the real FLAG gradient cycle 103 called a finding - when the strata are cut on the pair itself, and are flat within 0.73% and non-monotone when the strata are cut externally. Monotonicity across deciles is therefore NOT evidence against a coverage artefact. Cycle 103 is safe because it cut externally, The standing duplicate-check exits 126 and returns nothing at 3,000 files; it caused no duplicate ids, but the audit it should have been doing found two superseded entries that never name their successor - one of them a live safety verdict, The third silent-failure class, audited after being named in four king logs: exactly one of seven bigWig analyses is raw-coverage, the single incommensurable pair is already labelled three times, and the cross-pipeline ratio comparison was bounded at the point of measurement, The two silent-failure classes found in cycle 59 did not propagate through the corpus - 13 candidates, all legitimate on reading, and the audit corrected itself twice before it found anything, This graph states 624 falsifiers and records the outcome of 17 - a figure this entry first published as 8, because the detector did not know the corpus's own ALL-CAPS convention for recording one, This repo's "the pan-fusion core is 7.4-fold cleaner per unit signal than the singleton tail" divides one control column by a denominator summed over seven baits — on a per-bait denominator the ratio is 1.06 and p goes from 0.074 to 0.91, Three quarters of KMT2A::AFF1's top partners are shared with other MLL fusions — and the two analyses that said otherwise were both artifacts of how the source table was built, Two measurement defects confirmed: a denominator summed over seven baits made the pan-fusion core look 7.5x cleaner than it is (corrected fold 1.06, p 0.91), and my own gap report counts as coverage in the join that measures the gap, What the Polycomb beat searched, what it counted as absent, what it read and rejected, and the two absences it got wrong before hand-back, When a family shares a complex, the result is family-selective, not member-selective - the protein version of this graph's inhibitor-class rule
10. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    SWI/SNF panel lineage-split DepMap dependency scan, 28 genes

    Oncofusion knowledge graph, analyses/swi-snf-lineage-split/run.py · DepMap Chronos gene effect via Open Targets GraphQL 2026
    used by "SWI/SNF" is three different machines: cBAF, PBAF and ncBAF - and only one of them looks KMT2A-selective, 4,589 genes with no prior, and menin comes first (run), A gene-set permutation is anti-conservative for a protein complex, and the fix is to permute the other margin, A null that costs one command, run first rather than afterwards - and SSRP1 is genuinely LESS required in KMT2A-rearranged lines, while the rest of the chaperone layer is not much of anything, AF10/MLLT10 fusions drive aberrant self-renewal of non-stem progenitors — found only by opening the paper, AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing, AFF2 (FMR2) — the SEC-L2 scaffold, and the AFF-family member with no dependency anywhere, AFF3 (LAF4) — the AFF-family scaffold of SEC-L3, and a KMT2A fusion partner in its own right, AFF4 — the textbook SEC scaffold that KMT2A::AFF1 cells need LESS than other leukemias, AFF4 has never been placed in a haematopoietic cell state — every SEC cell-state result is in another tissue, AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature, AFG2B / SPATA5L1 (Q9BVQ7) - the second 55LCC ATPase, the only read-queue member with an annotated nuclear pool, pan-essential in 95.5% of lines, and the one that touches PSIP1, ALDH1L2 (Q3SY69) - the second-largest fall in the serine/one-carbon panel (-3.40), named nowhere in this graph, zero of 1,258 DepMap lines require it, and a tumour SUPPRESSOR where it has been studied, ALL-PO (CVCL_1069) — an infant KMT2A::AFF1 line that is already in DepMap and has never been screened, AQR (Aquarius) - a spliceosomal RNA helicase whose loss impairs MLL-rearranged leukaemia stem cells, Are the menin partners outliers, or did we pick them? (run), ARID1A (BAF250A) - required in the MLL-AF9 screen, present at the Myc enhancers, and flat in every selectivity measure, ARID2 (BAF200) - the PBAF subunit whose loss ACCELERATES MLL-AF9 leukaemia, and which is nonetheless required to maintain it, ASH2L — the WRAD subunit that has DPY30's essentiality profile and none of its co-dependency signal, ASH2L is studied in leukaemia and has never been placed in a cell state there — and the KMT2A-rearranged arm needs it LESS than average, ASH2L: the only WRAD subunit ever perturbed in a KMT2A::AFF1 cell, once, weakly, ATF3 (P18847) - the one cascade member the DOT1L inhibitor hits HARDER than the menin inhibitor, and a transcriptional REPRESSOR being counted as a stress output, ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor, ATF5 / ATFX (Q9Y2D1) - the most reproducible repression in the earlier dataset, and the three probe values the graph quotes are SEM, not RS4;11, ATXN7L3 — the shared adaptor that actually sets H2Bub levels, and a KMT2A-selectivity hit nobody in this field has looked at, BAF inhibition against the menin axis: on the evidence it is a complement, not an alternative - and the case rests on one laboratory, BARD1 (Q99728, 777 aa) - BRCA1's obligate partner, and the member of the cascade that actually touches the chromatin: its PxVxL motif at 569-573 binds HP1, and CTCF binds BARD1 rather than BRCA1 in an undamaged cell, BCL2 — the one co-dependency in this beat that already has an approved drug pointed at it, BCOR — the other side of the CBX8 conflict, and a target you would be inhibiting in the direction of disease, BCORL1 — the BCOR paralogue that is in the only crystal structure of the PRC1.1 core, and is the one PRC1.1 subunit nobody has tested in KMT2A-rearranged leukaemia, BICRA (GLTSCR1) - the ncBAF scaffold, zero PubMed records in KMT2A leukaemia, and a lineage-independent dependency in KMT2A-rearranged lines, BMI1 - the self-renewal gene that KMT2A fusions can do without, because HOXA9 covers for it at the same locus, BMI1, CDKN2A and CDK6 are one circuit, not three genes - and in KMT2A-rearranged disease it has a loop nobody has measured, BOD1L (BOD1L1) — the top SETD1A co-dependency, and the only candidate in this cycle tested by CRISPR tiling in a KMT2A::AFF1 line, Bounded absence: no published CRISPR or dependency screen in any infant-derived KMT2A::AFF1 model, BRCA1 (P38398, 1,863 aa) - named in 19 entries of this graph and never written: the HR mediator that CTCF and the H3K9 machinery are now reported to license at a break, and the only cascade member with a human therapy-related-leukaemia signal after topoisomerase II therapy, BRD4 — the best-drugged protein in this beat, and the one whose clinical result is already in and disappointing, BRD9 — dispensable in the average cell, rank 39 of 17,728 in KMT2A-rearranged lines, and with an on-target proof most epigenetic targets never get, BRD9 survives the null; CEBPA sits at the median of random genes, BRIP1 — downstream effector; rank 22 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, CALM-AF10 was meant to separate the menin axis from the HOXA programme - and the indirect evidence has now turned against separability, because KMT2A-free HOXA lesions keep proving menin-DEPENDENT, CBS (P35520) - the largest fall in the whole panel (-4.07), a token that collides with Cajal bodies AND CTCF binding sites, and a germline phenotype whose hallmark is thrombosis, CBX3 / HP1-gamma (Q13185, 183 aa) - the EUCHROMATIC HP1 paralogue, and the one CTCF actually prefers: the reader that carries BRCA1/BARD1 onto H3K9-methylated break chromatin, flat in DepMap, and a fusion partner in one KMT2A::AFDN AML patient, CBX5 / HP1-alpha (P45973, 191 aa) - the H3K9me3 reader that recruits the writer, the writer's destroyer and the second writer; and the one member of that circuit with a published pro-leukaemic role, as part of an MLL leukaemia stem-cell programme that immortalises myeloid progenitors WITHOUT Hoxa or Meis, CBX8 — the one place in this interactome where two good papers flatly disagree, and DepMap takes a side, CCND3 (cyclin D3) — the scan's rank-4 hit, and the only one of the four with a drug class already in children, CCNT1 / cyclin T1 — the regulatory half of P-TEFb, and the one part of the elongation core with any dependency variance, CCNT2 / cyclin T2 - expressed in every KMT2A-rearranged line and needed by none of them, so the cyclin T split is real rather than an expression artefact, CDC73 / parafibromin — the PAF1C subunit that was genetically tested, and a second tumour suppressor required by the fusion, CDK12 - a different kinase with a different cyclin, and in KMT2A-rearranged lines it is an anti-dependency, CDK13 - rank 171 of 17,728, equally deep in both lineage halves, and essentially unstudied in leukaemia, CDK4 - the other half of the class name, and it is NOT the gene this disease needs, CDK4/6 inhibition in KMT2A-rearranged leukemia — the class was already tried, and the trial never reported, CDK6 - the kinase the KMT2A-r rationale actually rests on, and the pan-essentiality flag and the selectivity are both true, CDK9 - the kinase the whole SEC story runs through, which this graph held only as half of a two-protein summary entry, CDK9 inhibition in KMT2A-rearranged infant leukemia — the most direct hit on the hijacked machinery, and the narrowest window, CEBPA — ranks 25th on the KMT2A scan and sits at the median of random genes, Cell type or the fusion? MV4-11 LOOSENS the confound the graph's own lineage split was built on - but its founding paper calls it biphenotypic, CHAC1 / BOTCH (Q9BUX1) - the largest single effect in the whole menin-inhibitor response (-5.11 log2) and a glutathione-degrading pro-apoptotic enzyme, named nowhere in this graph until now, Co-dependency scan: what else does a menin-dependent cell need? (run, not planned), Compute: which backend for which piece of phase-2 work, and the one constraint that matters for each, CREBBP (CBP) — binds the KMT2A transactivation domain, which no fusion keeps, CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose, CTH (P32929) - gene_exact returns TWO reviewed records and the second is VSIG2, a third instance of rule 62 on the same symbol the Ensembl resolver already got wrong, CTNNB1 (beta-catenin) — the Wnt link inside DotCom, and the one piece of KMT2A biology that is not chromatin, CTR9 — a PAF1C subunit the field assumes is a fusion partner, with no paper that tested it, CXCL12 (SDF-1) - the marrow retention chemokine this graph already had a mechanism entry for and no entry for the protein, CXCR4 - named in eighteen entries of this graph with no entry of its own, and the one niche molecule with a drug, a paediatric trial and a direct KMT2A-r experiment, CXXC1 (CFP1) - a CpG-reading COMPASS subunit that binds MLL1 and recruits it to HoxA7, and whose CXXC domain cannot substitute for KMT2A's in a fusion, CXXC4 (IDAX) - TET2's CXXC domain, sitting on a separate gene after an evolutionary inversion, and annotated cytoplasmic only by similarity, CXXC5 (RINF) - the one CXXC reader measurably down-regulated in KMT2A-rearranged AML, which is the only direction-of-change any family member has in this disease, Dataset: DepMap CRISPR essentiality - the bulk download is now behind a captcha, and this changes how phase 2 must be built, DDIT3 / CHOP (P35638) - the cascade's middle node, and the symbol whose two reviewed human records and 16,893-record alias are the worst identifier trap in this beat, Decitabine - degrades DNMT1, and in ALL cells it degrades KMT2A too: the hypomethylating agent with two mechanisms and a menin-inhibitor combination already in the clinic, Decitabine's target is a pan-essential gene, and KMT2A-rearranged lines depend on it LESS than other leukaemia lines - rank 16,949 of 17,728, DEK - a chromatin architectural protein that gatekeeps H3.3 and boosts HP1alpha, is a fusion partner in its own right, and is the substrate behind this graph's third CK2 route, DepMap across the whole CTCF-BRCA1 licensing cascade in one call: CTCF is a near-total pan-essential at -1.83, BARD1 is deeper than BRCA1, the HP1 paralogue that matters is flat while the one that does not is four times deeper - and DNA2's row has 80 screens where everything else has 1,258, DepMap cannot answer whether MEN1 is selectively required in NUP98-rearranged lines, because it contains none — and it takes only ~3 to answer it, DepMap co-dependency cannot rank the 97 unwritten interactome partners either, pan-cancer OR restricted to leukaemia lines: mean r 0.008 rising to 0.023 under restriction while the leukaemia-matched null rises with it, giving 1.08x the chance-expected count. Recomputing the null on the restricted lines is the only thing that stopped this being reported as a discovery, and my first pass criterion called a below-chance result a PASS, DNA2 (P51530, 1,060 aa) - EXO1's redundant partner in long-range resection, dual mitochondrial and nuclear, and the gene whose DepMap row is measured in 80 screens where every other cascade member has 1,258, DNMT1 - the antagonist: its CXXC domain reads unmethylated CpG in order to STOP methylation, and it is the only heterologous CXXC that runs an MLL-AF9 leukaemia, Do the other WRAD subunits behave like DPY30? (run, not planned), Does a menin inhibitor create a BMI1 dependency? A two-arm knockout-under-drug experiment in KMT2A::AFF1 cells, with the CDKN2A readout that decides the mechanism, Does the menin co-dependency survive lineage adjustment? (run), DOT1L — the target where excellent chemistry met an indirect interaction, DPF2 (BAF45D) - a cBAF-specific reader required in MLL-AF9 cells, with one PubMed record in this disease and no selectivity signal, DPY30 — the smallest COMPASS subunit, and the only one that behaves like a selective dependency, DPY30 aims the H3K4 machinery while ASH2L powers it, which is why only DPY30 tracks menin, EAF1 — the interactor with an actual subnuclear phenotype: an MLL fusion evicts it from Cajal bodies, and the bodies themselves fall apart, EAF2 (U19) — the EAF paralogue whose ELL-binding domain is a mapped tumour-suppressor module, described entirely in prostate cancer, EHMT2 / G9a - the H3K9me1/2 writer that sets whether the KMT2A LOCUS ITSELF rearranges, and whose loss delays MLL-AF9 AML in a supplementary figure nobody can find, EIF2AK4 / GCN2 (Q9P2K8) - the amino-acid-sensing eIF2alpha kinase: transcript flat, and NOTHING has ever been published about it in any KMT2A-rearranged system under either name, ELL — rank 7 on selectivity and the deepest common-essential in the top ten, with one NMR fragment to its name, ELL has never been placed in a cell state — its entire leukaemia state literature is its own fusion transcript used as a residual-disease marker, ELL2 — a SEC subunit that almost no cell needs, next to a paralogue almost every cell needs, ELL3 — the third ELL paralogue, which marks enhancers and recruits SEC rather than sitting inside it, Endogenous aldehydes and silenced ALDH2 — the rival explanation for an FA dependency in acute leukaemia, and the only one with an experiment behind it, EP300 — the acetyltransferase whose paralogue split is the finding: p300 is the 108th most KMT2A-selective gene, CBP is the 16,950th, EPC1 - the NuA4/TIP60 subunit whose C-terminal fragment grips MBTD1, making it the contact a leukaemic fusion cannot do without and the mildest dependency in that whole complex, Erasing H2Bub: three deubiquitinases share two adaptors, and removing the enzyme the field names lowers the mark instead of raising it, ERG - a heptad stem-cell transcription factor whose 15 mentions in this graph are mostly about OTHER diseases, and whose best-established role in B-ALL is loss of function, not cooperation, Every age-restricted target nominated by the model is pan-essential or invisible, Every gene in the H2B ubiquitination axis, ranked against the whole genome for KMT2A selectivity — and the writers are not selective, Exactly one gene in 8,248 beats menin, and it is FLT3, EXO1 (Q9UQ84, 846 aa) - the long-range resection nuclease at the bottom of the CTCF cascade, and the point where this graph's two independent 2026 HR papers meet: CtIP starts the resection, EXO1/DNA2 finish it, Expressed, required, maintains — three different claims about where a gene matters, and the graph must label which one it has, EZHIP (CXorf67, CATACOMB) - a protein that jams PRC2 by pretending to be an H3K27M oncohistone, and the 3' payload of MBTD1's other fusion, FAAP100 — FA core complex; rank 36 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FACT (SPT16-SSRP1) - the chromatin arm of transcription elongation, and a clean negative result for the KMT2A::AFF1 interactome, FANCA — FA core complex; rank 20 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FANCD2 — the FA effector menin binds, and the weakest FA gene in this graph's own selectivity scan, FANCF — FA core complex; rank 11 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FANCG — FA core complex; rank 27 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FANCI — ID2 module; rank 15 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FANCL — FA core complex; rank 26 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, FBXL19 - a CpG-island reader that delivers CDK-Mediator and RNF20 to developmental promoters, and whose experimental localisation is mostly cytoplasm, Fifth explanation for the Fanconi signal: it runs through menin binding FANCD2, not crosslink repair — tested, refuted, prior 0.07, First systematic enumeration of the KMT2A fusion interactome, and the diff against this graph, Five new arginine-axis entries in one day, and not one of their genes is a selective dependency - in an instrument that recovers MEN1 at p=5e-05 and TDRD3 is not a dependency at all, FLT3 / CD135 — the strongest non-tether co-dependency, and a target the field may have written off on a badly executed trial, FLT3 is the second largest KMT2A-selective dependency and it is HEAVIER in B-lineage than in myeloid cells, Four ways a WDR5 result can be true and have nothing to do with KMT2A, FOXC1 - a mesenchymal forkhead factor derepressed in HOXA-high AML that blocks monocytic differentiation, with a real mechanism and zero published connection to KMT2A rearrangement, GAS7 — growth-arrest-specific F-BAR protein, the second pillar of the forced-oligomerisation model, GATA2 - the least KMT2A-selective gene on this beat, ranking 17,591 of 17,728, and a factor whose leukaemia biology is about stemness and chemoresistance rather than about the fusion, GFI1 — the repressor that LSD1 inhibitors actually work through, and the scan hit whose literature points the opposite way, GPHN (gephyrin) — synaptic scaffold and molybdenum-cofactor enzyme, KMT2A fusion partner in t(11;14)(q23;q24), GPT2 (Q8TD30) - the EARLIEST-moving and third-most menin-selective gene in the serine arm, and the one REH cannot control because it is below the array floor there, H2B ubiquitination is upstream of the fusion's defining mark and is still not a target, because the fusion uses a general mechanism rather than a special one, H2BK34ub — a second ubiquitination site on H2B that also stimulates DOT1L, written by MSL1/MSL2, and important at HOXA9 and MEIS1, HB11;19 (CVCL_8227) is KMT2A::MLLT1, not KMT2A::AFF1 — and it is the most menin-dependent line in the panel, HCFC1 (HCF-1) — a founding member of the wild-type KMT2A complex that the fusion cannot bind, HEXIM1 - the brake pad, the only part of the reservoir anyone has deliberately raised in a KMT2A-rearranged cell, and the result was that the leukaemia grew less, HEXIM2 - the paralogue that holds the bound P-TEFb pool constant when HEXIM1 is removed, which is why single-gene experiments on this axis read flat, How BAF maintains KMT2A-rearranged leukaemia: it holds the blast undifferentiated, and it does so in other leukaemias too, HOXA10 - named in 27 entries as part of the HOXA output and never written; a posterior HOX paralogue that is regulated with HOXA9 and tested far less often, HOXA7 - the one HOXA paralogue besides HOXA9 with a published loss-of-function requirement for MLL-fusion transformation, and the requirement was shown in vitro only, HOXA9 — required for survival of established MLL-rearranged leukemia, and dispensable for starting it, Hypothesis (first clause REFUTED 2026-09-21): the fusion's core dependencies may be cell-type-invariant, but the cell type itself is not fixed, ICE1 — the 2,266-residue scaffold of the little elongation complex, the subunit that acts at initiation, and zero papers connecting it to leukaemia of any kind, ICE2 — the fourth LEC subunit, the one with no assigned step, no structure, and no leukaemia literature, IDH1 - the same oncometabolite, the same assay, and NO enrichment in KMT2A-rearranged AML (6.7% vs 7.6%, p = 0.53): the control that makes the IDH2 result interpretable, IDH2 - the hypermethylating co-mutation that IS enriched in KMT2A-rearranged AML (16.0% vs 10.4%), at p = 0.0004 and not the p < 0.0001 the abstract claims, If HOXA9 substitutes for BMI1 at CDKN2A, then a menin inhibitor should CREATE a BMI1 dependency that does not exist at baseline, IGF2BP3 — an RNA-binding protein required for MLL-Af4 leukaemogenesis in mice that CRISPR barely touches in human lines, IGF2BP3 marks the fetal programme AND is required for B-ALL survival - the trap is not expression-for-requirement, it is requirement-for-state-maintenance, IKZF3 / Aiolos (Q9UKT9) - the B-lymphoid transcription factor that goes UP under both inhibitors, regulates BCL2, and is a degradable drug target, IMPDH2 - the rate-limiting guanine-nucleotide enzyme whose inhibition lowers LEDGF and menin protein and dismantles the MLL-fusion complex itself, In all 24 read-queue rows inspected, the candidate is the BAIT and this project's protein is the prey - nobody has ever tagged KMT2A, AFF4, menin or PSIP1 and found any of them; plus seven bounded rejections, ING2 - the PHD finger that reads H3K4me3 and turns it into REPRESSION, and whose founding co-crystal is filed under the MOUSE accession, Interactome round 4: what was counted, what was written, what was rejected, and the two alias traps this beat fell into and caught, Interactome round 5: 42 queries, 13 entries written, 9 candidates rejected with counts, and four alias failures of four different kinds, Is the selectivity just pan-essentiality? No: it explains 0.4% of the variance, ITGA4 (CD49d, the alpha-4 of VLA-4) - the niche adhesion subunit with an approved antibody and a 2026 KMT2A-r B-ALL xenograft result, JAZF1 - a 243-residue zinc-finger protein that turned out to be a NuA4 subunit, and the 5' donor of the commonest endometrial-stromal-sarcoma fusion, KANSL1 - the NSL scaffold, and the protein through which WDR5 binding to NSL and to MLL/COMPASS is mutually exclusive, KANSL2 - the second NSL subunit in the WDR5 structure, with one leukaemia record and a dependency profile that is the mirror image of a selective target, KAT7 / HBO1 — an enzyme, a leukaemia-stem-cell dependency, and an inhibitor already in a crystal, KAT8 (MOF) — a genuine KMT2A partner that binds the half of KMT2A the fusion discards, KAT8/MOF sets the quiescence set-point of the blood stem cell - and its catalytic activity is separately required for MLL-AF9 leukaemia, which DepMap also cannot see, KDM1A (LSD1) — acts at the loci the fusion binds, spares normal stem cells, and still has never been shown to touch the fusion protein, KDM2A - the CXXC reader that antagonises KMT2A leukaemogenesis, and displaces wild-type MLL rather than the fusion, KDM2B — 513 potent ChEMBL rows that are mostly patent bins, and no structure of the domain they claim to inhibit, KDM3B - the H3K9me1/2 demethylase that sits on the KMT2A breakpoint cluster region and keeps it from rearranging, and the only leukaemia gene in this graph whose direction of effect flips between the myeloid and the lymphoid literature, KDM4C (JMJD2C/GASC1) — recruited by the retained KMT2A N-terminus, drugged in a KMT2A::AFF1 line, and invisible to CRISPR, KDM5A / JARID1A / RBP2 (P29375, 1,690 aa) - already in this graph as the 3' half of NUP98::KDM5A and never written as a gene: the H3K4 demethylase CTCF recruits to a break, the only ERASER in the licensing cascade, and the one cascade member with no blood-lineage ChIP-seq anywhere, KDM6A / UTX — a commonly asserted KMT2A partner that belongs to a different complex and points the opposite way, KITLG (stem cell factor) - the niche ligand of KIT, zero mentions anywhere in this graph, and its own symbol collides with the SCF ubiquitin ligase this corpus talks about constantly, KMT2A rearrangement creates a requirement for the FA ubiquitin-transfer module — the claim, its prior, and the four results that would kill it, KMT2A-rearranged cells need the factors retained through elongation MORE and the factors ejected at pause release LESS — a six-gene asymmetry nobody has reported, KMT2A::ABI1 (MLL-ABI1), t(10;11)(p11.2;q23) — the partner that binds an elongation partner, KMT2A::AFF1 cells run with more free P-TEFb than other leukaemias - and if they do not, this graph's CDK9 story is about recruitment only, KMT2A::AFF1 needs a developmental window and a lymphoid-committed cell — the cell type is a condition of transformation, not a consequence of it, KMT2A::EPS15 (MLL-AF1p), t(1;11)(p32;q23) — the infant-ALL cytoplasmic partner, KMT2A::SEPTIN11 (MLL-SEPT11) — the fusion that cytogenetically impersonates t(4;11), KMT2A::SEPTIN5 (MLL-SEPT5 / MLL-hCDCrel-1), t(11;22)(q23;q11.2), KMT2B has no haematopoietic cell-state result, and two of the seven records that look like one are about a different gene in muscle, Knock down CDK13 and CDK12 side by side in lineage-matched KMT2A and non-KMT2A lines, and ask whether the DepMap signal is real, LARP7 - the clamp on the 3' end of 7SK, whose human loss-of-function phenotype is dwarfism and intellectual disability rather than cancer, LEDGF/p75 is required in the leukaemic cell and dispensable in ADULT MOUSE HAEMATOPOIESIS - not in the normal organism, which its germline knockout does not survive, LEO1 — the weakest-evidenced entry in the PAF1 complex, and the gene whose Ensembl id this project already got wrong once, LEPR - the marker that defines the SCF-supplying perivascular niche cell, the receptor for an adipocyte signal that blunts cytarabine, and a gene_exact query that returns the wrong protein first, Lineage-split dependency for the round 5 panel: 23 genes, 2 myeloid-identity artefacts caught, and 3 survivors of which one has no literature at all, Lineage-split DepMap: neither USP2 nor MAML2 is a dependency in KMT2A-rearranged lines, and USP22 is, Live DepMap on fourteen genes: CDK6's pan-essentiality is a lineage artefact of pooling, and not one gene of the homeodomain cofactor code is a dependency anywhere, LMO2 (rhombotin-2) - a bridging LIM-only protein, higher in pro-B and MLL-rearranged B-ALL, anti-selective in KMT2A-rearranged lines, and a physical partner of the KDM5A module that NUP98::KDM5A retains, MAML2 (mastermind-like 2) - the Notch co-activator that is a KMT2A partner in leukaemia and in thymoma, MBD1 - one protein with a methyl-CpG reader and three CXXC fingers, only the third of which binds unmethylated CpG, and whose CXXC fails the MLL-AF9 swap, MBTD1 - the H4K20me1/me2 reader inside NuA4/TIP60, the quiescence brake on adult haematopoietic stem cells, and the 3' partner of ZMYND11::MBTD1 whose methyl-lysine cage the fusion carries and does not use, MEAF6 - a 191-residue subunit shared between the NuA4/TIP60 and MOZ/MORF acetyltransferase complexes, whose only two UniProt localisations both cite papers that are not localisation studies, Measure replication stress directly in KMT2A-rearranged cells instead of inferring it from a dependency ranking, Measure the pausing index at KMT2A::AFF1 target genes — the one number the whole pausing argument turns on, and it has never been reported, MECOM (EVI1) - a cell-of-origin memory that MLL-AF9 maintains but cannot install, present in about 43% of MLL-rearranged AML and marking the chemoresistant half, MED12 - a named menin-associated protein that restrains menin's transcriptional output, and the anchor of the CDK8 module a KMT2A-r AML paper co-targets, MED23 — rank 16 of 17,728, survives the lineage split with the B-lineage half stronger, and is not absent from the leukaemia literature after all, MED23 confirmed, and the zero-record claim corrected: it is not literature-free in leukaemia, an shRNA screen already found it, MED26 — a common-essential Mediator subunit with a solved 87-residue domain and no ChEMBL target entity at all, MEF2C — an output-program transcription factor, the second-strongest leukemia co-dependency, and a gene humans cannot spare one copy of, MEIS1 — rate-limiting for MLL leukemia stem cells, a surviving co-dependency, and embryonic-lethal when removed, MEN1, DOT1L and LEDGF dependency in the only CALM-AF10 cell line DepMap has screened, Menin dependence varies 2.83x across KMT2A::AFF1 lines and the spread survives at three donors - but RS4;11 is last on KMT2A itself, so the heterogeneity is largely fusion-dependence heterogeneity, and DAXX is the one gene decoupled from it, Menin has never been tested against SET::NUP214 - a bounded absence in three independent registers: literature, cell-line dependency data, and the founding paper's own blots, MEPCE - absent from this graph until now, and a deeper pan-dependency than BRD4, MLLT1 / ENL — the reader that beat its own paralogue, with a degrader already in hand, MLLT11 / AF1q — a 90-residue protein with no chromatin domain and no oligomerisation module, KMT2A partner in t(1;11)(q21;q23), MLLT3 / AF9 — a famous fusion partner that is not a dependency, and a hub domain that is, MLLT3 runs normal stem-cell self-renewal and is required in essentially no leukaemia line — the inverted case, MLLT6 (AF17) — the fourth KMT2A fusion partner inside DotCom, and the only one no cell needs, MOF/KAT8 is required by both KMT2A::MLLT3 and NUP98::HOXA9 — which is why it is the wrong kind of convergence, MSL1 - the MSL scaffold that carries MOF and MSL2 to HOXA9 and MEIS1, with zero records against KMT2A and a mapped, drawable domain architecture, MSL2 — the second H2B ubiquitin ligase, sitting inside the MOF complex the KMT2A fragment contacts, MTHFD1 (P11586) - the inverse case: the deepest leukaemia dependency in the serine/one-carbon panel, and the drug does not touch it, MTHFD2 (P13995) - the one panel member with a published MLL-AF9 in-vivo survival result, and the serine gene that behaves exactly like an ISR gene, MV4-11 (CVCL_0064) — KMT2A::AFF1, yes, but a ten-year-old's acute monocytic leukaemia with FLT3-ITD, MYB — rank 17 of 17,728 for KMT2A selectivity, and the only node in this beat where partial, transient suppression eradicated leukaemia without stopping normal blood production, MYC - named in a hundred entries of this graph and, until now, written in none, MYO1F — unconventional myosin IF, a KMT2A partner with neither a transactivation domain nor a dimerisation domain, NELF — the brake that has to come off for elongation, absent from the KMT2A literature, and less required in KMT2A-rearranged lines than in any other leukaemia, NELFB (COBRA1) — the NELF scaffold, and the single least KMT2A-selective gene of the sixty least selective in the genome, NELFE (RDBP) — the RNA-binding subunit of the pausing brake, and the alias that inflated a record count fifty-fold before it was caught, None of the five selective unwritten interactors survives a null, and FANCD2 - the one that carried a story - is the second weakest of them at p = 0.12, Nothing beats menin after all, and the statistic that said otherwise ranks narrowness, NPM1 (nucleophosmin) — the nucleolar shuttle whose mutant form binds chromatin with KMT2A and buys the second menin-inhibitor indication, NSD1 - the H3K36 dimethyltransferase whose catalytic activity a NUP98 fusion cannot do without, and which cells can otherwise lose almost for free, NSL complex - the MOF assembly that five NUP98 fusions bind alongside MLL1, and that competes with MLL/COMPASS for WDR5, NUP214 / CAN - the cytoplasmic-face nucleoporin that CRM1 docks against, and the 3' partner of four leukaemia fusions, NUP98 - the mobile nucleoporin that leaves the pore to put H3K4me3 on haematopoietic promoters, and the shared 5' donor of seven fusions in this graph, Of 262 curated interactors this graph never wrote down, five are selectively required in KMT2A-rearranged lines - and one of them is FANCD2, which binds menin and sits inside a selectivity signal this project has failed four times to explain, P-TEFb (CDK9 + cyclin T1) — the deepest dependency and the clearest therapeutic-index wall in the interactome, PAF1 — pan-essential subunit of a complex the fusion needs, where the window is in the interface and not in the protein, PAF1C couples transcription to H2B ubiquitination - but the RTF1 half of that coupling is yeast-specific and contested in human cells, and PAF1C is one of the few complexes that genuinely binds the retained KMT2A fragment, PAF1c is needed by the normal stem cell and by the leukaemia, and the genes it regulates in each are different, Pan-essentiality: checked for all eight, and cleared — but a shallow dependency is its own problem, PBX3 - the TALE cofactor the field says is non-redundant, and the experiment that says HOXA9 is the redundant one, PCGF1 — the PRC1.1 scaffold whose LOSS causes menin-inhibitor resistance, so it is a restraint on the leukaemia rather than a dependency of it, PES1 / pescadillo (O00541) - the missing PeBoW subunit of this graph's nucleolar block, pan-essential in 96.9% of DepMap lines, and the one myeloid experiment on it shows proliferation WITHOUT a differentiation block, Phase-2 dossier: CBX8 - the candidate this review would kill, and the reasons are all numeric, Phase-2 dossier: DOT1L - perfect chemistry, a completed paediatric trial, and a target that has already had its answer, Phase-2 dossier: KAT7/HBO1 - the only enzyme here with a real active site, a co-crystal inhibitor and a co-dependency that survives lineage adjustment almost intact, Phase-2 dossier: the ENL YEATS pocket - two co-crystals, nanomolar chemistry, and a co-dependency that only appears inside leukaemia, Phase-2 dossier: the LEDGF/PSIP1 IBD - the best therapeutic index in the interactome, a solved structure, two hotspot residues, and three compounds in the world, Phase-2 dossier: the menin MBM1 pocket - the only candidate here that is already de-risked, and therefore the only one whose gate is clinical, PHF1 (PCL1) - the PRC2 accessory subunit whose Tudor domain reads H3K36me3, and the 3' acceptor shared by three different NuA4-donor fusions in two different tumour types, PHGDH (O43175) - the largest menin-SELECTIVE fall in the whole serine panel, and the least essential gene in it, PICALM (CALM) - the clathrin-assembly adaptor that is the 5' partner of CALM::AF10, PICALM::MLLT10 (CALM-AF10), t(10;11)(p13;q14) - the DOT1L-dependent HOXA fusion that contains no KMT2A, PLK1 — nominated as an infant-restricted vulnerability, and one of the most essential genes in the genome, PPIE (CyP33) — the off-switch the fusion throws away, and the reason PHD3 reinsertion kills transformation, PRC1.1 — the complex this graph has been leaning on, now enumerated subunit by subunit, and the enumeration changes the conclusion, PRMT1 — an MLL-fusion cofactor whose first-in-human inhibitor was terminated for lack of efficacy and toxicity, PRMT3 - the mostly-cytoplasmic ribosome methyltransferase that also writes the ACTIVATING mark on the residue PRMT5 represses, has an acute-leukaemia degrader, and is invisible to this project's own mark-query rule, PRMT5 - the type II arginine methyltransferase this graph has reached for sixteen times, whose mark on H3R2 RECRUITS WDR5 to chromatin, PRMT6 - the enzyme whose mark is MUTUALLY EXCLUSIVE with the MLL complex on the H3 tail, and which shuts HOX genes down by evicting WDR5, PRMT7 - the enzyme whose class is still disputed, which represses a KMT2 enzyme by writing H4R3me2s, which is required for leukaemia stem cells in CML, and whose only H3R2 claim is the one this beat was dispatched to check, PSAT1 (Q9Y617) - a DIRECT ATF4 target, and the gene that shows the brief's menin-selectivity claim was PHGDH's and not the pathway's, PSIP1 / LEDGF p75 — the tether with the best therapeutic index in the whole interactome, PSPH (P78330) - the third serine-synthesis enzyme, high in Ewing sarcoma alongside PHGDH and PSAT1, and the one that does NOT fall in leukaemia, Put the same fusion in two cell types and degrade it, rather than cutting the gene, RAD50 (Q92878, 1312 aa) - the third MRN subunit, and the one that settles whether the complex is incoherent in DepMap: mean -0.114 with 5.2% of 1,257 lines below -0.5, so it sits with NBN and not with MRE11, yet it covaries with MRE11 across cell lines as tightly as BRCA1 does with BARD1, Ranking every KMT2A::AFF1 partner on dependency versus essentiality, RBBP5 — a WRAD subunit that binds the one part of KMT2A every fusion throws away, REST / NRSF - the transcription factor that delivers G9a to chromatin through MED19 and MED26, is a paediatric cancer gene, and is the worst search token in this corpus, RNF2 (RING1B) — the one catalytic subunit shared by every PRC1 flavour, which is why the genetics point two ways at once, RNF20 (BRE1A) — the H2B ubiquitin ligase that MLL-rearranged leukaemia needs, and that DepMap says every other cancer line needs just as much, RNF40 (BRE1B) — the obligate partner of RNF20, and the one of the pair with a documented link to differentiation and bivalency resolution, RSBN1 / KDM9 (Q5VWQ0) - an H4K20me2/me3 demethylase in the AFF4 proximity record, and the only read-queue partner that is a chromatin enzyme at all, RTF1 — the PAF1C subunit the graph was missing, on a complex that does bind the retained KMT2A fragment, Run the menin inhibitor against CALM-AF10 next to a KMT2A-rearranged control - the missing two-week experiment, Running the HOXA programme is not enough to need menin, RUNX1 — activated by KMT2A::AFF1 and then bound by the reciprocal fusion, in the one published two-oncoprotein circuit in t(4;11), Scoring complexes instead of genes, and the controls disagree with each other (run), SELE (E-selectin) - the vascular-niche chemoresistance receptor with a clinical antagonist, zero KMT2A papers, and a paediatric trial that a company restructuring closed, Separate ncBAF from canonical BAF in KMT2A-rearranged cells: paired degron, paralogue-pair CRISPR, and complex-resolved CUT&RUN, SEPTIN11 (septin-11) — KMT2A fusion partner at 4q21, adjacent to AFF1, SEPTIN2 (septin-2 / NEDD5) — KMT2A fusion partner in t(2;11)(q37;q23), SEPTIN5 (septin-5 / CDCrel-1) — KMT2A fusion partner in t(11;22)(q23;q11.2), SEPTIN6 (septin-6) — cytoskeletal GTPase, KMT2A fusion partner in t(X;11)(q24;q23), SEPTIN9 (septin-9 / MSF) — cytoskeletal GTPase, KMT2A fusion partner in t(11;17)(q23;q25), Sequencing phase 2: the order to do things in, ranked by information per unit cost, SESN2 / sestrin-2 (P58004) - the ISR's exit into mTORC1: a leucine sensor that falls 1.4 log2 under either inhibitor, and the one panel member with a real structural record, SETD1A — a leukemia dependency whose enzyme activity turns out to be dispensable, SETD1B - the COMPASS paralogue round 4 deferred for want of a leukaemia paper, which had already been published: its SET domain sets H3K4me3 breadth and MYC in MLL-rearranged AML, SETDB1 (Q15047, KMT1E/ESET) - the second H3K9me3 writer: it represses the fusion's own targets via the PAF1 complex, it is required in fetal liver HSPCs and in leukaemic stem cells, and it is a broad dependency in half the CRISPR panel - three published directions that do not agree, SGF29 — a SAGA reader with a Tudor pocket, intermediate essentiality, and the best unexploited profile in this beat, SHMT2 (P34897) - the mitochondrial one-carbon step, and the paralogue that rises while it falls is why its loss may not matter, SIAH1 — the E3 ligase that degrades AFF1, and that the reciprocal AF4::KMT2A fusion escapes, SIAH2 — the second AFF1 E3 ligase, reported only alongside SIAH1 and never separated from it, SIRT1 - not a fusion partner, the least required gene in KMT2A-rearranged lines, and the one target here where the direction of the drug is genuinely disputed, Six countermeasures, no head-to-head — so rank their targets: CDK6 is 13th and CDK4 is 16,052nd, Six times the null pool, and the criterion I had been using turns out to be the problem (run), SKP1 — a pan-essential housekeeping adaptor that is also a structural subunit of PRC1.1, which makes it the one PRC1.1 member that can never be a target, SMARCA2 (BRM) - the paralogue nobody measured, and the reason a DepMap rank cannot predict what a BRG1/BRM inhibitor does, SMARCA4 (BRG1) - a tumour suppressor by mutation and a maintenance requirement by knockdown, and in KMT2A-rearranged lines the dependency data say it is not selective at all, SMARCB1 (BAF47/INI1/SNF5) - zero entries, zero mentions and, on the evidence, correctly so: it is lost in primary AML rather than required, SMARCD1 (BAF60A) - rank 59 of 17,728, survives the lineage split, and had never been named in this graph or studied in this disease, SMARCD2 (BAF60B) - named in one KMT2A paper in the whole of PubMed, and the dependency it carries is myeloid identity, not the rearrangement, SPI1 / PU.1 — a myeloid dependency that a KMT2A panel mistakes for its own, SPP1 (osteopontin) - the endosteal matrix protein whose knockdown prolongs survival in MLL-AF9 mice, and the one KMT2A::AFF1-specific paper about it cannot be read, SS18 - a SWI/SNF subunit that is itself a fusion partner in another cancer, and a lineage-independent dependency in KMT2A-rearranged lines, SSRP1 - the small FACT subunit, and the only protein in this beat that binds a confirmed KMT2A-fusion interactor directly (LEDGF/p75, via HMG box to PWWP), STC2 / stanniocalcin-2 (O76061) - the second largest effect in the signature (-4.40 log2), a SECRETED glycoprotein, and its localisation rests on curator inference alone, SUPT16H (SPT16) - the large FACT subunit: pan-essential, an anti-dependency in KMT2A-rearranged lines, and drugged in MLL-r leukaemia by a compound that works through p53, SUPT4H1 (SPT4) — zero papers linking it to KMT2A, and the 138th most KMT2A-selective gene out of 17,728, SUPT5H (SPT5) — the pausing factor the KMT2A field has never studied, and the counted absence that undercuts 'the fusion hijacks pause release', SUV39H1 (O43463, KMT1A) - the H3K9me3 writer: a menin effector whose binding region OVERLAPS the drug pocket at its N-terminus, a tumour suppressor in MLL-AF9 AML, recruited to double-strand breaks by CTCF, and never localised at the KMT2A locus in any cell type, SWI/SNF enumeration for KMT2A-rearranged leukaemia: every subunit searched, every count dated, and what was read and rejected, SWI/SNF split by sub-complex and by lineage: every KMT2A-selective subunit is in the ncBAF module and none is in canonical BAF, TET1 (LCX / CXXC6) - a CXXC reader that is also a KMT2A fusion partner, and whose leukaemic fusion throws its own CXXC domain away, TET2 is NOT a CXXC protein - its CXXC domain is a separate gene, and it is DEPLETED in KMT2A-rearranged AML rather than enriched, TET3 - the CXXC reader whose domain prefers 5-carboxylcytosine, and the only CXXC-family gene ever found mutated in a KMT2A-rearranged cohort, The 82.9% population scored twice as many defect-shaped notes as the same-source one, and reading six of them says the comparison is confounded by our own DepMap numbers, The comparator arm was never screened for pathway-equivalent lesions — and it made us understate our own result, The confounder every previous analysis flagged and none removed, removed (run), The deferred KMT2A partners: what the counts actually are, what was written, what was rejected, and three numbers the previous enumeration got wrong, The dependencies this project keeps failing to see are the buffered ones: paralogue redundancy and guide placement, not absence of biology, The easy null was not the problem, and nothing in 2,018 genes beat menin, The ELL/EAF activation module (ELL2, EAF1, EAF2) — a general elongation stimulator, not a KMT2A-fusion dependency, The entire experimental model space for SET::NUP214 is two cell lines, neither of them in DepMap, with no mouse model found - so every mechanism in this beat is n = 1 or n = 2, The FA core complex does not move together, and that is the strongest argument against reading this as a Fanconi-pathway dependency at all, The Fanconi signal is in both lineage halves, survives the correlation-proof null, and is still there when lineage is held constant, The fusion may reach the H2B ubiquitin mark through MOF-MSL rather than through RNF20/RNF40 — the one configuration in which this layer could be selective, The histone/ribosomal shape in the read queue is three classes, not one - H2BC21 is distinguishable AND essential at -0.70, so neither existing rejection ground reaches it, and HMGA1 is a harder dependency than MEN1 or KMT2A, The LEDGF PWWP–H3K36me nucleosome contact: how the fusion learns its address, The lineage confound, tested: the FA signal is in both halves, the B-lineage half is stronger, and it survives holding lineage constant, The little elongation complex (ELL + ICE1 + ICE2 + ZC3H8) — the other place ELL works, and why ELL's pan-essentiality is probably not about the fusion, The LSD1–GFI1 SNAG interface — a second worked example of drugging an undruggable repressor through its partner, The menin-FANCD2 bridge is dead: it was tested in 2008, the mutation spectra differ, and there is no genetic interaction, The MRN asymmetry adjudicated: RAD50 -0.114 sits with NBN -0.047 and not with MRE11 -0.542, so it is not one subunit against two - and yet MRE11 and RAD50 covary across 1,208 cell lines at r = 0.554, the same as BRCA1 against BARD1, against a background of zero, The phase-2 go/no-go register: the numbers we agree to before we look, The serine/one-carbon arm is NOT a pure ATF4 output: it is menin-selective where the ISR arm is compound-independent (median margin -0.55 vs -0.02, p=0.011), and it is ABSENT in KMT2A::MLLT3 AML where the ISR arm is present, The serine/one-carbon programme has a SECOND, chromatin-proximal menin input that is not ATF4 - which is why PHGDH is menin-selective where CHAC1 is not, and why the arm is absent in KMT2A::MLLT3, The six open read-queue members split three rejected and three earning entries - and the flattest dependency in the set, ASB7, turns out to be a negative regulator of H3K9me3 via SUV39H1 degradation, The stricter null, and it partly reverses the last one (run), The SWI/SNF dependency in KMT2A-rearranged leukaemia belongs to ncBAF, not to canonical BAF - and the drug in the clinic is aimed at the wrong module, The transcriptional CDK with a real KMT2A-rearranged selectivity signal is CDK13, not CDK9 - and nobody has looked, The whole protein-coding genome: menin is still first, and a metabolic pathway appears (run), The within-donor null: MEN1 sits at the 99.4th percentile of variation between two sublines of ONE child, and at the 98.9th between donors - menin dependence is not heterogeneous between patients so much as unstable in this background, TOP3B (O95985) - the only human topoisomerase that works on RNA as well as DNA, delivered to the MYC promoter by an arginine mark, and it has poisons, TRIB3 / TRB3 (Q96RU7) - a pseudokinase that is the ISR's own brake, falls 3.1 log2 under either inhibitor, and has a published physical link to the MLL complex in the wrong tissue, TRRAP — a 3,859-residue DotCom subunit that every cell needs and no KMT2A cell needs more, Two of the three top new dependencies are myeloid identity, not KMT2A, Two published ways a fusion oncoprotein engages BAF - be a subunit, or recruit it with a disordered domain - and KMT2A has never been tested for either, UBE2A (RAD6A) — the E2 for H2B ubiquitination in one primary account, contradicted by another, and with no KMT2A literature at all, UBE2T — E2; rank 33 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all, UROS: searched and empty — the scan's rank-15 hit has no leukemia literature, and no entry was written, USP2 (ubiquitin carboxyl-terminal hydrolase 2) - the ninth commonest KMT2A partner and the only one that is a deubiquitinase, USP22 — a KMT2A-rearranged dependency that works through MEIS1 protein stability, and is not an interactor of the fusion, USP8 (UBPY) - the second deubiquitinase KMT2A partner, three patients, same UCH logic, VCAM1 - the stromal counter-receptor for VLA-4, and in MLL-fusion AML its UPREGULATION is anti-leukaemic, which is the opposite of the expected sign, Verdict: 'eleven cell lines, none infant-derived' is right about the screen and wrong about the material — at least ten infant KMT2A::AFF1 lines exist, WAC — the adaptor that hands RNF20/RNF40 to RNA polymerase II, and the only member of this axis whose knockdown alone killed MLL-rearranged ALL cells, WDR5 — the partner that is not a partner: it binds the half of KMT2A the fusion throws away, WDR5 is partitioned, not shared: NSL and SET1/MLL complexes compete for one pool, so perturbing NSL changes how much WDR5 reaches the wild-type KMT2A complex the fusion depends on, WDR5 maintains a differentiation block — in CEBPA p30 AML, not in KMT2A-rearranged disease, and a title check nearly buried it, WDR5: an established wild-type KMT2A contact, and no fusion contact ever measured, What this beat looked at and did not write: rejected candidates, with counts for both halves of every search, Which leukemia co-dependencies survive uncertainty? (run, with p-values), WRAD reaches KMT2A fusions only as a passenger on a recruited wild-type complex, YWHAG / 14-3-3 gamma (P61981) - the only cross-laboratory reproduced KMT2A proximity partner in the curated record, and UniProt places it nowhere near the nucleus, ZEB2 — an EMT transcription factor that turns out to be a leukemia dependency and a guardian of lineage fidelity, ZEB2 level sets how easily a KMT2A-rearranged blast can change lineage under CD19-directed pressure, ZFP64 — rank 4 on selectivity, and there is not even a ChEMBL entry for it
11. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    ENCODE files ENCFF271ZAZ and ENCFF951BQB (ENCSR000BRT, CBX3 K562) and ENCFF457RFT and ENCFF668JZI (ENCSR000EDB, BRCA1 HeLa-S3), GRCh38

    ENCODE Project Consortium · ENCODE Portal 2026
    used by "Zero called peaks" is a statement about a reproducibility threshold, not about whether the protein was there - and in this graph the difference was worth a whole cycle's conclusion, All 39 bounded absences in this graph, classified by WHAT BOUNDS THEM: 23 rest on an abstract-level literature search, 7 on a threshold or a query filter, and only 2 on reading the primary source
12. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    absence-no-resection-factor-localised-at-the-kmt2a-locus, data.ABSENCE_CLASS_CHANGED_cycle_63

    Q35-encode-reanalysis; verified by the king · this graph 2026
    used by "Zero called peaks" is a statement about a reproducibility threshold, not about whether the protein was there - and in this graph the difference was worth a whole cycle's conclusion
13. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
    analyses/the-histone-mark-floor-was-on-disk/run.py and results.json, this repository, computed 2026-09-23 with zero downloads from bins in C:/Temp/q125 and C:/Temp/q117; replicate identity established in analyses/what-the-deposit-says-its-channels-are; GSE299100 comparison figures from analyses/the-same-condition-noise-floor/results.json

    Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
    used by [CORRECTED cycle 110 - F2 not F1, and the floor spans ~20x across arms] Measuring the same-condition histone-mark noise floor from replicates already binned on local disk, zero downloads: 15.97% at the peaks and 3.03% at D1, so the borrowed 10% floor was 3.3x too generous in the bulk - independently reproducing, in a different deposit, cell system, mark and antibody, what the GSE299100 beat measured at 9.4% and 2.4%
14. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    analyses/what-a-share-does-to-a-known-collapse/run.py and results.json, this repository, computed 2026-09-23 from the GSE149183 series supplementary sitepro tables; noise floors from analyses/the-histone-mark-floor-was-on-disk; cycle 103 profile from analyses/where-does-the-displaced-fusion-share-go/results.json

    Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
    used by [CORRECTED cycle 117 - the columns were not scaled] [CORRECTED cycle 110 - the gradient is NOT an artefact] Grading this project's own stratified-share statistic on data where the truth is known: it is not blind to a global collapse, it is DISTORTING - a pure 5.84x genome-wide loss with no redistribution whatsoever is rendered as a monotone gradient rising from the peaks into the sparse deciles, reporting a 9.70x INCREASE at D10 where the absolute change is 1.66x, and 62 of 66 strata are confidently outside the measured noise floor, [CORRECTED cycle 117 - the columns were not scaled] Correcting yesterday's correction: share(s) = absolute(s)/G is an exact identity, so a within-sample share reproduces the true SHAPE faithfully and misstates only the LEVEL - but that is the worse failure, because 34 of 66 strata lost up to 68% of their signal while reading as gains of up to +440%
15. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.computation_on_this_corpus
    Population and field-multiplicity audit of data.absence_class across the Oncofusion corpus

    king, Oncofusion knowledge graph, cycle 68 · scripts/check-absence-class.py 2026
    used by `absence_class` is one token per ENTRY while the entries carry seven to eleven distinct gaps each, so it is a summary label - and inside gene-nbn it hid a verifiably false ENCODE claim for three cycles, I audited whether a valid absence token is the RIGHT token, expecting the gene-nbn error to be widespread. It is not: 3 of 4 read were correctly scoped, the scoping is done in prose, and 39 of 40 flagged entries carry no machine-readable scope, Only 20% of this corpus's absences carry a machine-readable query, which I nearly published as an auditability crisis - and then read five of the other 80% and found every one bounded by a better instrument, The absence-class checker globbed a filename instead of the field, so it walked 40 of 243 entries and reported them as the whole population - and the 40 it called invalid turn out to be two independently re-invented vocabularies, The unbiased half of the token audit: 0 of 15 never-measured tokens are wrong, the entries my filters could NOT see are the best-scoped ones, and two scoping fields already existed that I had claimed to add
16. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    ENCODE Portal: /targets/NBN-human/ HTTP 200 and 4 released TF ChIP-seq experiments for target.label NBN

    ENCODE Project Consortium · ENCODE Portal 2026
    used by `absence_class` is one token per ENTRY while the entries carry seven to eleven distinct gaps each, so it is a summary label - and inside gene-nbn it hid a verifiably false ENCODE claim for three cycles
17. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    PubMed E-utilities esearch, KMT2A-sarcoma literature size measurements

    National Center for Biotechnology Information · PubMed E-utilities 2026
    used by 0.2% of profiled sarcomas, median age 38, and every number here has a denominator that is not the population, AF10/MLLT10 fusions drive aberrant self-renewal of non-stem progenitors — found only by opening the paper, AFF4 has never been placed in a haematopoietic cell state — every SEC cell-state result is in another tissue, ANP32E - the eviction chaperone: the only protein in this beat whose job is to take a histone variant OUT of chromatin, ASF1B - the paralogue that actually does the work in blood: it occupies >70% of H3.3 nucleosomes in fetal liver while HIRA is dispensable there, ASH2L is studied in leukaemia and has never been placed in a cell state there — and the KMT2A-rearranged arm needs it LESS than average, ASH2L: the only WRAD subunit ever perturbed in a KMT2A::AFF1 cell, once, weakly, Ask whether the childhood AML in the KDM3B-haploinsufficiency syndrome cohort carried a KMT2A rearrangement - the one human test this mechanism has, and it is a records request, ATF3 (P18847) - the one cascade member the DOT1L inhibitor hits HARDER than the menin inhibitor, and a transcriptional REPRESSOR being counted as a stress output, ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor, ATF5 / ATFX (Q9Y2D1) - the most reproducible repression in the earlier dataset, and the three probe values the graph quotes are SEM, not RS4;11, ATRX - DAXX's partner and the other arm of H3.3 deposition; in this disease its only signal is in the four lymphoid cell lines, CHAC1 / BOTCH (Q9BUX1) - the largest single effect in the whole menin-inhibitor response (-5.11 log2) and a glutathione-degrading pro-apoptotic enzyme, named nowhere in this graph until now, CHAF1A (CAF-1 p150) - the subunit that reads PCNA and makes histone deposition replication-coupled, and the half of CAF-1 that leukaemia has not looked at, DAXX - the strongest of five selective dependencies among 254 unwritten interactors but it FAILS Bonferroni correction (p=0.0128, rank 4 of 254), and not one paper connects it to this disease, DDIT3 / CHOP (P35638) - the cascade's middle node, and the symbol whose two reviewed human records and 16,893-record alias are the worst identifier trap in this beat, DEK - a chromatin architectural protein that gatekeeps H3.3 and boosts HP1alpha, is a fusion partner in its own right, and is the substrate behind this graph's third CK2 route, Does a menin inhibitor create a BMI1 dependency? A two-arm knockout-under-drug experiment in KMT2A::AFF1 cells, with the CDKN2A readout that decides the mechanism, Does KMT2A::AFF1 change histone turnover at its target genes, or does it read a chromatin state it did not make?, DPY30: the founding negative does not cover it, and it has no KMT2A contact to lose, EHMT2 / G9a - the H3K9me1/2 writer that sets whether the KMT2A LOCUS ITSELF rearranges, and whose loss delays MLL-AF9 AML in a supplementary figure nobody can find, EIF2AK3 / PERK (Q9NZJ5) - the kinase whose transcript is flat under both inhibitors, whose arm IS functionally required in MLL-AF9 leukaemia stem cells, and whose common name collides with phospho-ERK, EIF2AK4 / GCN2 (Q9P2K8) - the amino-acid-sensing eIF2alpha kinase: transcript flat, and NOTHING has ever been published about it in any KMT2A-rearranged system under either name, ELL has never been placed in a cell state — its entire leukaemia state literature is its own fusion transcript used as a residual-disease marker, ENL is the highest-expressed target in cerebellum and nobody has studied it there, Enumerating the WRAD-versus-KMT2A-fusion literature: what was searched, found, rejected and could not be read, Every published KMT2A partner in sarcoma, counted - plus the papers this beat read and did not use, so the next beat does not repeat the search without the decisions, FLT3 is the second largest KMT2A-selective dependency and it is HEAVIER in B-lineage than in myeloid cells, Giving the reopening conditions an owner, and the first run changes two decisions (run), H2A.Z - two genes of identical length, one KMT2A paper, and the measured quantity is a ratio nobody has repeated, H3.3 - the graph's most-named epigenetic object with no entry, 325 mentions, and not one measurement in a KMT2A-rearranged cell, H3K27ac - named in 40 entries with no entry of its own, and the strongest KMT2A::AFF1 measurement of it is invisible to every abstract search, H3K4me2 - the LSD1 substrate, and the only histone mark ever compared head-to-head between two KMT2A fusion partners in human cells, H3K9me1 - the only H3K9 state whose level at the KMT2A breakpoint cluster region has ever been measured, and it is LOW there until KDM3B is removed, H3K9me2 - named sixty times in this graph with no entry of its own, the mark that defines the lamina-associated domains the fusion's targets sit beside, and the only H3K9 methylation state ever measured at a KMT2A fusion's own binding sites, H3K9me3 - measured in KMT2A::AFF1 cells, in a supplementary figure, and it rises GENERALLY where H3K27ac falls SPECIFICALLY in the same experiment, H4K16ac - the one mark rule 80 actually rescued in this beat: elevated at MLL-AF9 targets in a supplementary figure, and the reason MLL1's own enzyme turns out to be dispensable, HIRA - the chaperone that puts new H3.3 into transcribed gene bodies, and therefore the machine that runs this graph's clearance clock, Histone H3.3 - one protein, two genes, and the reason this graph should not file it as gene-h3f3a, IKZF3 / Aiolos (Q9UKT9) - the B-lymphoid transcription factor that goes UP under both inhibitors, regulates BCL2, and is a degradable drug target, ING2 - the PHD finger that reads H3K4me3 and turns it into REPRESSION, and whose founding co-crystal is filed under the MOUSE accession, JMJD1C / KDM3C - named in three entries of this graph as one of the only loci that loses fusion occupancy under a menin inhibitor, ranked 23rd of 17,728 for KMT2A-selective dependency in this project's own scan, with the strongest published shRNA effect in SEM - and no entry until now, KAT8/MOF sets the quiescence set-point of the blood stem cell - and its catalytic activity is separately required for MLL-AF9 leukaemia, which DepMap also cannot see, KDM3B - the H3K9me1/2 demethylase that sits on the KMT2A breakpoint cluster region and keeps it from rearranging, and the only leukaemia gene in this graph whose direction of effect flips between the myeloid and the lymphoid literature, KDM6B (JMJD3) - UTX's paralogue, which this graph has argued about for cycles without writing, KMT2B (MLL2/MLL4) - the other protein menin binds, and the one this graph never wrote, KMT2B has no haematopoietic cell-state result, and two of the seven records that look like one are about a different gene in muscle, LEDGF/p75 is required in the leukaemic cell and dispensable in ADULT MOUSE HAEMATOPOIESIS - not in the normal organism, which its germline knockout does not survive, MLL2 and MLL4 each name two different genes, and the ambiguity is in the papers, not in the databases, Niche biology HAS been measured in KMT2A-rearranged disease - but in the index fusion itself there is essentially one marrow-niche molecular experiment, from 2011, Niche proteins enumerated and NOT written in this tranche, and four symbol traps measured while doing it - three of them new to the registry, No H3K9me3 writer and no H3K9me3 reader has ever been localised at the KMT2A locus IN THE PUBLISHED LITERATURE, in any cell type: 20 queries, 5 alias spellings, 0 on-topic records - and CORRECTED 2026-09-22: the SUV39H1 half is no longer unanswerable from archives. ENCODE still holds zero SUV39H1 experiments (verified by three routes with controls), but GEO holds two deposits, and Q114 read both at this locus and found the writer ABSENT, No menin inhibitor has ever enrolled a KMT2A-rearranged solid tumour, and the one solid-tumour menin trial that read out was terminated for lack of efficacy, No proximity, degron or tiling-CRISPR study has ever been run on WRAD in KMT2A-fusion leukaemia, Nobody has ever asked whether a KMT2A-rearranged sarcoma needs menin: zero records, and the reason is that no model exists, One blot: phospho-eIF2alpha and ATF4 protein in RS4;11 under VTP-50469, with a dose-response and a DOT1L comparator arm, PAF1c is needed by the normal stem cell and by the leukaemia, and the genes it regulates in each are different, PPIL4 - a spliceosomal cyclophilin whose BioID neighbourhood is the widest overlap with this graph's subject matter of any bait, PPP1R15A / GADD34 (O75807) - the eIF2alpha phosphatase adaptor, and KMT2A fusion proteins were shown to BIND it and block its apoptosis in 1999, a paper this graph has never cited, PRMT3 - the mostly-cytoplasmic ribosome methyltransferase that also writes the ACTIVATING mark on the residue PRMT5 represses, has an acute-leukaemia degrader, and is invisible to this project's own mark-query rule, PRMT5 - the type II arginine methyltransferase this graph has reached for sixteen times, whose mark on H3R2 RECRUITS WDR5 to chromatin, PRMT6 - the enzyme whose mark is MUTUALLY EXCLUSIVE with the MLL complex on the H3 tail, and which shuts HOX genes down by evicting WDR5, RAD51 - the one repair protein measured AT the KMT2A translocation hotspot, and ATM controls how much of it loads there, RBBP5: the thinnest fusion literature of the four subunits, and one chromatin datum, RBM15 — m6A adaptor, AMKL fusion partner, and a named hit in the only KMT2A-biased RNA-protein CRISPR screen, Re-ran this corpus's literature bounds against the live index: 40 of 40 stored zeros still return zero, none has expired - and 19 stored 'queries' abbreviate their own clauses, so they were never runnable bounds at all, REST / NRSF - the transcription factor that delivers G9a to chromatin through MED19 and MED26, is a paediatric cancer gene, and is the worst search token in this corpus, Rule 80 measured: the abstract index missed the best KMT2A measurement for three marks out of seven, and confirmed the absence for the other four, SELE (E-selectin) - the vascular-niche chemoresistance receptor with a clinical antagonist, zero KMT2A papers, and a paediatric trial that a company restructuring closed, SEPTIN2 (septin-2 / NEDD5) — KMT2A fusion partner in t(2;11)(q37;q23), SESN2 / sestrin-2 (P58004) - the ISR's exit into mTORC1: a leucine sensor that falls 1.4 log2 under either inhibitor, and the one panel member with a real structural record, SET / TAF-I beta - a histone chaperone that binds the retained N-terminus of MLL, co-activates HOXA with it, and is itself a HOXA-activating fusion partner that recruits DOT1L, SGF29 — a SAGA reader with a Tudor pocket, intermediate essentiality, and the best unexploited profile in this beat, SGK1 / SGK (O00141) - a MENIN-SELECTIVE upregulated kinase, a druggable AGC family member, and the panel's worst platform disagreement, SKIC8 (formerly WDR61) — the last PAF1C subunit, and a gene-symbol trap worth recording, SMARCB1 (BAF47/INI1/SNF5) - zero entries, zero mentions and, on the evidence, correctly so: it is lost in primary AML rather than required, SMARCD1 (BAF60A) - rank 59 of 17,728, survives the lineage split, and had never been named in this graph or studied in this disease, SMC3 - the cohesin ATPase arm, the only subunit whose partial loss has been tested directly against a KMT2A fusion, and it was tolerated, SOCS2 (O14508) - a MENIN-INHIBITOR-SELECTIVE member of the signature, a cullin-5 E3 substrate receptor, and its likeliest upstream regulator is MEF2C rather than ATF4, SPI1 / PU.1 — a myeloid dependency that a KMT2A panel mistakes for its own, SRCAP - the other H2A.Z-loading motor, whose entire PubMed intersection with MLL is five neurodevelopmental gene panels, STC2 / stanniocalcin-2 (O76061) - the second largest effect in the signature (-4.40 log2), a SECRETED glycoprotein, and its localisation rests on curator inference alone, SUMO1 - not a complex partner of LEDGF but a modifier of it, and the interactome row cannot tell you which, SUPT6H (SPT6) - the histone chaperone that rides with RNA polymerase II, which puts it closer to the fusion's proposed mechanism than anything else in this beat, The clearance clock has machinery, and the machinery says the clock does NOT stop in a non-dividing cell - a postmitotic heart replaces its histones with a half-life of about two weeks, The druggable node in the homeodomain code is the MEIS1-PBX3 interface, not the HOX-PBX one - and it is the only one nobody has attacked, The H3K9me1/2-CTCF brake: a KDM3B-G9a balance keeps the KMT2A breakpoint cluster region methyl-poor and CTCF-bound, and losing either end makes the locus amplify and rearrange - in epithelium, in primary cord-blood progenitors and in mice, The KMT2-family beat's own ledger: every query run, every gene dropped, every paper read and not used, The leukaemia term hid the answer for DPY30 — a rule-77 demonstration with clean rule-67 controls, TRIB3 / TRB3 (Q96RU7) - a pseudokinase that is the ISR's own brake, falls 3.1 log2 under either inhibitor, and has a published physical link to the MLL complex in the wrong tissue, UNC0642 - the G9a/GLP chemical probe that abolished doxorubicin-induced Kmt2a copy gain in mice, and the one intervention in this beat that has been given to an animal, USP1: an inhibitor, activity in B-ALL, and zero papers in the disease we care about, WDR5 maintains a differentiation block — in CEBPA p30 AML, not in KMT2A-rearranged disease, and a title check nearly buried it, WDR5: an established wild-type KMT2A contact, and no fusion contact ever measured, What the CTCF-protects paper actually cites: it names Gothe and Atkin in ONE direction-free sentence and declares the direction unresolved, and the two papers that ran the reciprocal experiment are absent from all 81 of its references
18. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    Proteins API proteomics-PTM, P51825 (AFF1_HUMAN): 36 modified residues, 22 phosphorylation sites including S588

    · EMBL-EBI 2026
    used by 14-3-3 theta binds AF4 at Ser588 and rides into the MLL-AF4 complex - the phospho-gated contact the fusion really does carry
19. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
    Oncofusion knowledge graph, protein-aff1-domains: AFF1 junction and exon-4 codon ranges

    · Oncofusion knowledge graph (internal) 2026
    used by 14-3-3 theta binds AF4 at Ser588 and rides into the MLL-AF4 complex - the phospho-gated contact the fusion really does carry
20. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
    Evidence redundancy audit of the Oncofusion knowledge graph

    · Oncofusion knowledge graph (internal analysis) 2026
    used by 18% of this graph's evidence is the same finding said again
21. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    q153 neighbour census over the 19,404-promoter window table, hg38 refGene

    Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
    used by 21.6% of this project's 19,404 promoter windows contain another gene's TSS, so an attribution caveat is corpus-wide and not an IRX2 peculiarity, A promoter window is not a gene: what it means to say a protein 'sits on gene X's switch', The IRX2 / IRX2-DT geometry re-derived from three annotations: 215 bp and 89.2% under RefSeq, 437 bp and 78.2% under Ensembl, and the verdict does not move
22. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    q153 per-entry promoter-window sharing census over this graph's gene and protein entries

    Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
    used by 21.6% of this project's 19,404 promoter windows contain another gene's TSS, so an attribution caveat is corpus-wide and not an IRX2 peculiarity, AFF1's promoter window is 92.5% AFF1-AS1's - worse than IRX2's 89.2% - and the project's promoter table has no row for the neighbour at all
23. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    analyses/q148-what-the-hoxa-negative-blast-occupies - results148_pcdhg.json; author interval list from p148752_152985_Menin_Signal.csv

    Q148 (Oncofusion Knowledge Graph) · this repository 2026
    used by 22 of the 37 promoters in the published B-ALL-specific peak class are the clustered protocadherin gamma locus - one 23.8 kb interval at chr5:141,473,710-141,497,491 - and it reads below an average promoter in all 27 samples, A uniform permutation null and a coverability-matched one disagreed on 18 of 23 samples for the same gene set - the null, not the data, decided the verdict, A73051 has the lowest oncoprotein share of the 25 untreated samples, and its sequencing depth is mid-range (37th-39th percentile), so the reading is not a depth artefact - the coordinator's cycle-135 qualification misidentified enrichment as coverage and was reversed in cycle 138, Hypothesis: the HOXA-negative, IRX1-negative KMT2A-rearranged blast is a low-oncoprotein-dosage state, not a state whose oncoprotein has been redirected to another programme, My own pre-registered reagent gate was the wrong test: it demanded a within-sample ordering from a class defined by a between-sample difference, and it failed by 2.5%, Spike-in-normalised paired KMT2A N- and C-terminal CUT&RUN, with a fusion-junction readout, in a HOXA-negative KMT2A-rearranged blast beside a HOXA-positive one from the same institution and batch, TB920488, the other HOXA-negative IRX1-negative blast, has the highest GMP-over-B-lineage ratio of any B-lineage sample in the deposit - and it fails both the matched null and the N-over-C test, The AML-over-ALL retention ratio is 93.5% denominator: the GMP-like numerator barely moves across 23 samples while the B-lineage denominator varies fourteenfold, so the ratio measures loss of the B-lineage class and not gain of the GMP-like one, The GMP-like peak class is fusion-enriched in exactly one of GSE252378's 27 KMT2A-rearranged samples - p318, the sample it was derived from - once the null is matched on how coverable those promoters generically are, The GMP-like promoter class carries no N-over-C excess in 25 of GSE252378's 27 samples: whatever occupies it is seen equally by the fusion-side and the wild-type-side antibody, The GMP-like state is being used as a therapeutic anchor in two independent arguments, and its chromatin definition rests on one post-lineage-switch sample that also defined it, The GMP-over-B-lineage occupancy contrast tracks the sample's own lineage - and institution, which is the identical split in this deposit - and not its HOXA status, The pre-registered three-branch test of 'the HOXA-negative blast's oncoprotein is on the GMP programme' returns NEITHER/UNINFORMATIVE, on four independent gates, The published peak classes do transfer from 20 kb peaks to 2 kb promoters - the GMP-over-B-lineage ratio swings 8.56-fold across the pair that defined them - but only in the between-sample direction the classes were built in
24. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    Clustering the graph's 1,731 cited papers into research groups by author overlap, and recounting independence per entry

    · Oncofusion knowledge graph (internal analysis) 2026
    used by 3,288 papers are 2,025 research groups, the per-entry overcount is 11.7% and barely moved while the graph doubled - and TWO entries now rest on three papers from a single laboratory, where the first run of this analysis found none, Cell state in KMT2A-rearranged leukaemia is organised by lineage, not by the fusion — five independent axes split the same way, and every one of them splits AML from ALL rather than partner from partner
25. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    MeSH topic coherence per clustered group, as a test for two laboratories sharing a surname and initial

    · Oncofusion knowledge graph (internal analysis) 2026
    used by 3,288 papers are 2,025 research groups, the per-entry overcount is 11.7% and barely moved while the graph doubled - and TWO entries now rest on three papers from a single laboratory, where the first run of this analysis found none
26. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    analyses/is-discussed-the-same-as-draftable/run.py and results.json, this repository; run 2026-09-23 against the corpus as committed at git HEAD; population from analyses/core-complex-interactome-gap/results.json

    Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
    used by 43 of the 62 discussed-but-unwritten interactome partners (69%) are genuinely draftable from the committed corpus, so cycle 93's decision to work them before the 34 never-read ones has a real cost basis - but the clause that decided it was the citation clause I thought was safe, not the proxy I flagged as weak, and the 12 RNA-processing failures turn out to be a promiscuity census that should NOT be promoted partner by partner
27. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
    q153 enrichment of shared-window promoters among occupancy calls, GSE252378

    Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
    used by 43.9% of every occupancy call this project's rule makes is at a shared window it cannot attribute - twice the 21.6% base rate, in 27 of 27 samples, The 2.03-fold over-representation of shared windows among fusion-occupied promoters is generic promoter activity, not a fusion preference - and if it is not, the rule is finding something different from what it claims
28. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    q153 promoter-call discriminability over overlapping window pairs, GSE252378

    Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
    used by 43.9% of every occupancy call this project's rule makes is at a shared window it cannot attribute - twice the 21.6% base rate, in 27 of 27 samples, Attribution is asymmetric: q151's IRX2 negative survives a window that is 89.2% someone else's gene, and a positive in the same window would not have, PAF1 and MED29 return the identical occupancy call in all 27 GSE252378 samples, because 86.5% of their promoter windows is the same DNA, The 2.03-fold over-representation of shared windows among fusion-occupied promoters is generic promoter activity, not a fusion preference - and if it is not, the rule is finding something different from what it claims, When a 2 kb promoter call fires at a window shared with another gene, the neighbour fires too 91.3% of the time - against 4.3% for random promoter pairs
29. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41743263
    Intra-subtype heterogeneity shapes treatment response in KMT2A-rearranged ALL across all age groups.

    Hartmann AM, Bastian L, et al. · Hemasphere 2026
    used by 465 KMT2A-r B-ALL cases from 1 month to 89 years: partner predicts MRD clearance independently of age, KMT2A::MLLT1 is high-risk in childhood AML and the best-outcome cell in childhood T-ALL: the partner effect changes sign with lineage, The age confound: a partner arm and an age arm are nearly the same arm, The analysis that would settle it: a partner-by-lineage interaction term in pooled individual-patient data, Which partner, in which patient: KMT2A partner frequencies across six age-lineage strata
30. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    RCSB PDB search API, chemical-component query over UniProt-mapped entries

    · RCSB Protein Data Bank 2026
    used by 49 of 91 proteins in this graph have never had a drug-like molecule bound to them, What was searched for a structure and what came back empty: the KMT2A interface layer enumerated by accession pair, with the graph's own unmapped interfaces listed
31. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Proteins API proteomics-PTM, Q03164: PRIDE/PTMeXchange re-analysed post-translational modification sites

    · EMBL-EBI 2026
    used by 56 of KMT2A's 130 observed phosphosites survive the translocation, and the retained fragment is phospho-denser than the part that is lost, ATR phosphorylates KMT2A at S516 to break the SCF(Skp2) contact - a phospho-gated degradation switch sitting inside the retained fragment, CK2 phosphorylates KMT2A at T2724/S2726 to license taspase1 cleavage - and that site is in the half the fusion throws away, Cks1 and Cks2 hold both halves of KMT2A and set its level through the cell cycle, Every MS-evidenced KMT2A modification site, flagged retained or lost against the breakpoint - and the retained portion is NOT enriched for acetylation once you use the right null, KMT2A is a 14-3-3 client by proxy: the contact is on an AFF-family scaffold, not on KMT2A, KMT2A K1130 acetylation - the clock-controlled, SIRT1-erased site that every KMT2A fusion retains, 17 residues from the CXXC and 208 from the breakpoint, KMT2A K1130 ubiquitination - the same retained lysine as the clock acetylation site, carrying a chemically exclusive second mark that no KMT2A paper has ever mentioned, KMT2A K1133 acetylation - the better-evidenced half of the pair, and the one that appears in the CBP/p300 acetylome itself, KMT2A K1235 acetylation - inside the SIRT1-interaction region, retained, and the one member of the Choudhary quartet the source paper named and then never assayed, KMT2A K1364 SUMO - the one modification site inside the breakpoint cluster, retained in some patients' fusion proteins and lost in others, KMT2A K2528 SUMO2 - the single best-supported modification site on the whole protein, in the N320 fragment, and LOST in every fusion, KMT2A K636 acetylation - the forgotten member of the Choudhary quartet, retained, and outside the SIRT1-interaction region, Looking for a second handle on the retained fragment, and not finding one, Q25 enumeration: what was searched, found, rejected and could not be read across the five 3' partners, S821 is not phosphorylated in any dataset we could query: the 14-3-3 motif candidate is dead as stated, The partner, not the site, is what could make a retained KMT2A modification fusion-specific - the AML1-ETO mechanism asked of KMT2A::AFF1, The PTM-database route runs: K1130 and K1133 acetylation are independently re-observed in deposited mass spectrometry, K1133 in the CBP/p300 acetylome itself - and K1130 is also a ubiquitination site nobody has mentioned, The second handle on the retained fragment is CK2-dependent, not 14-3-3-dependent
32. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    citation-title-overlap-distribution: scoring every stored citation title against PubMed, and the census of citations that could not be scored at all

    · Oncofusion knowledge graph (internal analysis) 2026
    used by 577 citations carried a bare PMID with no title, so the audit built to catch a fabricated title could not see them: the third dropped denominator in this project
33. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    RCSB PDB entry 9XYC: Pol II-DSIF-SPT6-PAF1c-TFIIS-IWS1-ELOF1-LEDGF-nucleosome, cryo-EM 3.5 A, released 2026-03-18

    · RCSB Protein Data Bank 2026
    used by 577 citations carried a bare PMID with no title, so the audit built to catch a fabricated title could not see them: the third dropped denominator in this project
34. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    Noncoding gap ledger and absence-claim audit, computed over this graph

    · Oncofusion knowledge graph, analyses/noncoding-gap-ledger 2026
    used by 78% of the noncoding beat rests on a claim that nobody has done something
35. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    IntAct PSICQUIC query: detmethod:"MI:1314" (proximity-dependent biotin identification) intersected with the UniProt accessions of the super elongation complex, KMT2A, MEN1 and DOT1L, retrieved 2026-09-19

    · IntAct molecular interaction database (EMBL-EBI) 2026
    used by A 109-transcription-factor BioID network puts KMT2A next to MYC, YY1 and the ETS and KLF families - and AF10 next to ERG and KLF5, AFF1-NPM1 proximity: nucleophosmin is the only bait in any published proximity experiment whose neighbourhood contains AFF1, KMT2A and menin together, AGGF1 - an angiogenic G-patch protein whose BioID neighbourhood contains ELL, KMT2A, LEDGF, WDR5, KAT7 and RNA polymerase II, GPKOW - a G-patch spliceosomal protein whose BioID neighbourhood contains AFF4, ELL, ENL, CDK9, KMT2A and DOT1L, HNRNPC - the nascent-RNA packaging protein whose BioID neighbourhood contains AFF4, ELL, ENL, CDK9, cyclin T1 and KMT2A, PPIL4 - a spliceosomal cyclophilin whose BioID neighbourhood is the widest overlap with this graph's subject matter of any bait, Proximity-interactome beat: every query run, every paper rejected, and the one resource not reachable, Re-adjudicating the standing weak spot: proximity labelling HAS sampled AFF1, KMT2A and five leukaemia fusions - just never a KMT2A fusion, SMNDC1/SPF30 - a Tudor-domain splicing factor recovering ELL, ENL, KMT2A, PAF1 and KAT7, The 50-RBP BioID2 proximity map in K562 - the first modern proximity dataset this graph can reach, and AFF1 is in it as prey, The super elongation complex, sampled from the RNA side: ELL, ENL, CDK9 and cyclin T1 are all in the proximity record - ELL2, EAF1/2, AF9 and AF10 are at zero, Wild-type KMT2A is one of the most promiscuously labelled proteins in the proximity record - 25 of 50 baits - and menin is one of the least, at 3, ZRANB2 - the only bait in any published proximity experiment that reaches AFF1, AFF4, ELL, MLLT1, CCNT1, KMT2A and DOT1L at once
36. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.preprintPMID 41959134
    Combined Menin and XPO1 inhibition drive synergistic antileukemic activity in KMT2Ar and NPM1-m AML.

    Uddin MH, Dhiman S, Han Y, et al. · bioRxiv 2026
    used by A 2026 preprint puts XPO1 upstream of the menin interface itself: XPO1 reported to stabilise menin on chromatin and menin's binding to KMT2A and KMT2A::MLLT3, Chromatin-bound XPO1 as a shared, druggable address for HOXA-driven leukaemia - and the objection that it may be a general chromatin cofactor rather than an address at all, Has anyone tested an XPO1 inhibitor in a KMT2A-rearranged system? Yes - four preclinical studies and zero clinical ones, counted with the queries and the controls, Three measurements the export axis needs and nobody has published: XPO1 occupancy in KMT2A-r, H3K79 under an XPO1 inhibitor, and CRM1 occupancy under a menin inhibitor, XPO1 / exportin-1 (CRM1) - the cell's main nuclear export receptor, and a chromatin-bound transcriptional scaffold
37. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.reviewPMID 41481253
    A Unified Model: Chromatin-Bound Multicomponent Condensates.

    Lau BK, Haarer EL, Ahn JH, et al. · Cancer Research 2026
    used by A 2026 preprint puts XPO1 upstream of the menin interface itself: XPO1 reported to stabilise menin on chromatin and menin's binding to KMT2A and KMT2A::MLLT3, Docking onto pre-bound XPO1 is a general route to leukaemic chromatin, used by at least seven different lesions - not three coincidences
38. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    HGNC record HGNC:4015 (FUT4)

    HUGO Gene Nomenclature Committee · HGNC 2026
    used by A CD number names a reagent cluster, not a molecule - and four different things can sit behind one, CD15 (Lewis-x, SSEA-1, 3-fucosyl-N-acetyllactosamine) - a sugar, not a protein: the most-named antigen in this graph that no gene encodes, CD65 - a neutral glycolipid epitope (V3FucnLc8Cer), no gene, and the CD number that was split in two because its antibodies did not agree, FUT4 - the myeloid alpha-1,3-fucosyltransferase, a hub gene of the KMT2A-rearranged expression module, and the gene HGNC once officially called CD15, Q48 enumeration: every antigen, identifier and query this beat ran - including what was searched and rejected, and the three identifier traps found live
39. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB reviewed entry Q5QGZ9 (CL12A_HUMAN)

    UniProt Consortium · UniProtKB 2026
    used by A CD number names a reagent cluster, not a molecule - and four different things can sit behind one, CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce, Q48 enumeration: every antigen, identifier and query this beat ran - including what was searched and rejected, and the three identifier traps found live
40. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB reviewed entry P01903 (DRA_HUMAN)

    UniProt Consortium · UniProtKB 2026
    used by A CD number names a reagent cluster, not a molecule - and four different things can sit behind one, HLA-DR - a two-chain MHC class II heterodimer, so no single gene symbol denotes the surface marker a flow report calls HLA-DR, HLA-DRA - the non-polymorphic alpha chain, and the only HLA-DR chain any source on this beat names explicitly
41. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    Q125 re-analysis of GSE260742: the PCGF1-knockout genotype comparison at the eviction quartiles, read together with Q124's committed ARM-2 results. analyses/q125-the-h2ak119ub-arm-of-the-pcgf1-knockout/run.py, deviation1_hg38_container.py, deviation2_quartiles_and_conventions.py, deviation3_matched_null.py and results.json, run 2026-09-23

    · Iqbal Lab oncofusion knowledge graph 2026
    used by A cell that cannot deposit H2AK119ub is still evicted normally by a menin inhibitor and silences its vacated sites less - so a fusion-occupancy pharmacodynamic marker reads FULL ENGAGEMENT in exactly the genotype that responds worst, and this is now the fourth independent reason this corpus distrusts its own candidate markers, A ChIP-seq series of 158 samples that names no antibody anywhere in its metadata: the string 'antibod' occurs ZERO times in GSE260742's 419,824-character SOFT dump, and the reagents are in a supplementary spreadsheet the full text never quotes, A DOT1L inhibitor DOUBLES H3K79me2's mid-occupancy share (1.89-2.01) and a menin inhibitor CUTS it (0.69-0.78), in THREE murine samples one of which is a replicate (corrected cycle 97; the id still says four) with no overlap between the ranges, while the FLAG and MLL1-N channels sit at 0.905-0.929 under both drugs - so the two drugs do not act on one axis and 'a fraction of the achievable effect' presumes they do, A menin inhibitor's chromatin signature is organised by how much fusion it removed from each site, so every mark's response is a GRADIENT in eviction depth - and any readout that averages over a fusion-occupancy set is underpowered by construction, sometimes to exactly 1.000, A PCGF1-null cell - a published resistance genotype - loses H3K79me2 at the most-evicted sites exactly as a control cell does (0.6123 against 0.5860), so the new stratified readout scores it as fully drug-engaged, and stratification does NOT close the resistance blind spot it was hoped to close, A plateau criterion of the form 'the last interval contributes at most 10% of the total' fires when the last interval is NEGATIVE, so it reports a plateau for a series that reversed - and the guard that caught it was a separate monotonicity check written for a different reason, A Spearman over four quartiles has 24 orderings and 11 attainable values: rho = +1.000 is 1/24 = 0.042, |rho| = 1.000 is 2/24 = 0.083, and a +-20% band around 0.80 contains exactly one attainable value - so a perfect monotone over four strata is a weak statistic and must never be scored alone, Across eight murine contrasts and two human stratifications, the H3K79me2 eviction span separates a menin inhibitor from a DOT1L inhibitor 22-fold with no overlap AND detects menin engagement in a human cell whose aggregate reads 0.998 - two properties the aggregate has one of and one of, Across every same-condition H2AK119ub replicate pair in GSE260742 - nine pairs, three arms, both pairing shapes - rank concordance over the shared bin universe runs 0.6481 to 0.8796 and never reaches 0.88. The floors this thread scores against are all computed from pairs that are only moderately concordant., Across four VTP timepoints the within-sample share at a fixed top-0.1% set ranks perfectly inversely to the track's non-zero bin count (Spearman -1.000 over a 2.2-fold spread), so this project's standard instrument is confounded with ChIP enrichment - and the confound cannot be removed from inside one sample, All four input samples explicitly titled to the MLLAF9 ChIP arm of GSE260742 carry supplementary_file = NONE, so no input-referenced peak call is possible from the processed deposit - and the only deposited batch-1 inputs are two BAMs filed under a different arm's title with the same cell line and the same F3 filename prefix, All four murine VTP-50469 H3K79me2 contrasts in GSE260742 give Spearman -1.000 against eviction depth, and all four SGC0946 contrasts are flat: the sign reproduces four times out of four, the span only to within a factor of 1.99, and the replicate pair that establishes this was already binned on disk from cycle 82, Asked for the first time whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor: all three pre-registered criteria fire, and the set they fire on is 52 low-amplitude bins reaching 6x the genome median, 16 of them with no gene within 20 kb and 1 of 52 on a CpG island where 66% of the fusion's real sites are, At the 2 kb bins a menin inhibitor strips the most fusion from, both repressive marks GAIN and the active mark is LOST, monotonically in eviction depth (Spearman +1.000, +1.000 and -1.000) against a baseline-matched null that is flat to within 0.007 - the silencing clause of hyp-q117's prediction 5, measured for the first time, Bin the deposited H2AK119ub NTvKO 2 x 3 design: twelve files, three replicate conventions, public since 2025-11-23 and never analysed - it is the only route to the silencing clause of hyp-q117's prediction 5, it is the internal validation of the PCGF1 knockout this project's null depends on, and it carries a spread the H3K79me2 arm cannot, Both datasets this beat used carry an exposure mismatch that could have produced its results, and both were measured rather than argued away: MOLM13's MENIN pair is 4 days against everything else's 3, and GSE260742's VTP arms are 48 h against 72 h DMSO controls - the first moves a criterion by 0.024 and the second is matched between the two arms being compared, Cycle 107's headline same-condition floor of 3.03% at D1 is a median over a MIXED population of two genotypes. Restricted to the non-targeting control genotype it is 10.49%, so the borrowed 10% floor cycle 107 declared 'too generous by 3.3x' was in fact about right there - cycle 107's own F2, not its F1., Download and bin GSE127507's nine RS4;11 tracks (1.99 GB), rebuild a human binner, and score the eviction gradient in the index fusion KMT2A::AFF1 - the one measurement that would move this finding from KMT2A::MLLT3 to the disease this graph is about, Every chromatin readout of a menin inhibitor measured in this corpus is graded by how much fusion the drug removed from a site, so a within-sample average over the fusion's occupancy set is the wrong statistic and will report near-null in a responding cell - the signal is the SLOPE against eviction depth, not the level, Every noise floor in this thread is computed from within-sample shares, which sum to one. A purely local change therefore leaks into every other stratum by arithmetic, so no floor measured this way can distinguish 'the bulk moved' from 'the peaks moved and the shares renormalised'. A tighter floor is not thereby a cleaner one., Four matched input tracks, used for the two things an input is actually for: no chromosome-scale copy-number difference between the CRISPR clone and its control (max 1.0988, min 0.9470 in both processing states), and the knockout's H2AK119ub depletion from CpG islands is REAL because its own genomic DNA is enriched there, not depleted, GSE260742 declares an H3K27me3 second channel in its MLLAF9 Control arm, and no floor can be computed from it: all three rep1/rep2 pairs agree about only 86.31-88.90% of the shallower track's covered bins, so 11-14% of the SHALLOWER track's coverage falls where the DEEPER one has none. Depth cannot explain that., GSE260742's murine H3K79me2 samples come from FOUR distinct cultures - F3, F3_NT, NT and the PCGF1-KO clone C1 - and the titles name only two of the distinctions; 'Timecourse Control' is the F3 line, and 'NTvKO Control' covers two different filename families that are not the same material, GSE260742's PCGF1-knockout samples carry no deposited measurement of PCGF1 - no protein, no transcript, no sequencing of the edited locus - and the knockout rests on one guide and one clone with no rescue, so every null result from that arm is a null in an unverified genotype, GSE260742's VTP-50469 kinetic series carries no SGC0946 sample at any timepoint, so the denominator of F cannot be measured inside that batch and the fraction cannot be recomputed at any time - the only SGC0946 arms in the series are in different experiments at a single 72-hour or 7-day exposure, GSM8123377 and GSM8123403 carry the same supplementary filename but for the `.filt.bam` marker, yet GEO titles one 'MLLAF9 Control - DMSO H2AK119ub rep2' and the other 'NTvKO Control - DMSO H2AK119ub paired'. Scored directly, they are DIFFERENT libraries (Spearman 0.83), so the collision is a naming accident and both arm titles stand., GSM8123386/387/388 carry the GEO title 'MLLAF9 Control - ... H3K79me2 rep2' and the supplementary filename 'ChIP_K79_F3_NT_*'. Two committed analyses treat them as a separate culture and one asserts 'FOUR INDEPENDENT MURINE CULTURES'. If the titles are right there are three, one of them replicated. REPORTED AS A DISPUTE, NOT APPLIED., GSM8123405 is mm10 DATA in an hg38-DECLARED TDF - the exact mirror of Q122's mm10-declared / hg38-coordinate defect, in the same series - and the container silently clipped 72.8 Mb (2.76%) of chr13-chr19 and chrX at deposition; the declared extents name the container and max_end names the data, H3K4me3 is the one mark in GSE260742 with no declared second channel at all - six samples, no replicate suffix anywhere - so no same-condition pairing exists and its noise floor remains a transfer from another mark whatever else is measured., H3K79me2 reports only the DOT1L arm of a menin inhibitor's action and is blind to its PRC1.1 arm - so the mark's response is unchanged in a PCGF1 knockout even though PCGF1 loss is a published resistance mechanism, and a cell resistant through PRC1.1 loss would look fully drug-engaged by this readout, hyp-q117's prediction 5 is stored TWICE in one entry, in sections[0].body_md and in data.predictions[4], and the two strings differ - here is the scored outcome and the ALL-CAPS cashed text proposed for both locations, with the clause-by-clause verdict a one-line 'confirmed' would have hidden, hyp-q117's prediction 5 says 'in a PCGF1 OR BCOR knockout' and GSE260742's only genetic perturbation is the PCGF1 knockout: across all 158 sample records the strings BCOR, BCORL1, KDM2B, RING1 and RNF2 appear zero times as a perturbation, so a verdict on PCGF1 is not a verdict on the prediction as written, Hypothesis: the same-condition floor measured this way is largely a property of where the EXTERNAL stratifier puts its top bins relative to each library's own coverage, not of how well the two replicates agree - which would explain a twentyfold arm span sitting on top of rank concordances that barely differ., In a PCGF1-knockout clone the menin inhibitor still evicts and the vacated sites are still silenced, but LESS: H2AK119ub at the most-evicted quartile falls from 1.4450 to 1.1325 and from 1.0151 to 0.8842 in two replicate conventions, and H3K27me3 from 2.0007 to 1.0134 - so prediction 5's silencing clause is directionally right and quantitatively overstated, In human MOLM13, DOT1L's own occupancy under a menin inhibitor runs 0.1364 / 0.3083 / 0.3555 / 0.3913 across MLL1-N eviction-depth quartiles at Spearman -1.000 - the writer is lost in proportion to eviction depth, and its mark follows with a 2.9-fold smaller span, In MOLM13, human KMT2A::MLLT3, from the VTP-50469 founding paper's own series GSE127507, H3K79me2's aggregate response at MLL1-N's top sites is 0.998 - and the same sites split by eviction depth run 0.7746 / 1.0601 / 1.0995 / 1.1421 at Spearman -1.000, with the INPUT stratified identically staying flat, In murine MLL-AF9, H3K79me2's response to a menin inhibitor runs 0.586 / 0.819 / 0.928 / 1.187 across the four eviction-depth quartiles of the fusion's own top-0.1% sites - Spearman -1.000, span 0.6005, the largest of any mark measured in these bins except H3K27me3 - while the aggregate over the same 1,228 bins is a single unremarkable 0.8156, In murine MLL-AF9, knocking out PCGF1 changes the menin inhibitor's effect on H3K79me2 by 0.0166 of a share ratio and the DOT1L inhibitor's by 0.0101 - both smaller than the 0.0584 two control cultures of the same line already differ by, so the mark's drug response runs outside PRC1.1, In the only system where the fusion can be measured separately, a menin inhibitor removes MLL1 from wild-type-dominated sites harder than it removes the tagged fusion from its own sites - 0.1069 against 0.1793 - which is an on-target effect on normal KMT2A-menin complexes and an argument about toxicity nobody in this project has made from occupancy data, In the VTP-50469 timecourse Q124 declared uninterpretable, the eviction-gradient SPAN grows monotonically 0.104 / 0.198 / 0.214 / 0.335 at 4/8/16/24 h with Spearman -1.000 at every timepoint - while the aggregate it was suppressed for is non-monotone and perfectly anti-correlated with library complexity. The span is better behaved, and it is NOT clean., Menin inhibition removes the fusion from its strongest sites without relocating it to a new programme, and PRC1.1 fills in only where the eviction is complete - so the drug's chromatin action is subtraction plus local substitution, not redistribution, No washout, withdrawal, release or drug-removal sample exists in any of GSE260742's 158 records, so the third clause of hyp-q117's prediction 5 - that the programme returns on drug withdrawal - cannot be scored from the deposit that prediction names, and closing it needs a wet experiment rather than a download, Pre-registration protects against moving the goalposts and does not protect against this: a blacklist added to fix an unrelated artefact moved a pre-registered count from 42 to 52 across a threshold of 50, flipping the verdict from FAIL to PASS without anyone touching the threshold, Prediction 5 of hyp-q117 contains three separable claims - normal eviction, no silencing, and return on withdrawal - and the deposit it names supports one of them only through a clause it never states; the eviction and silencing clauses need tracks the NTvKO arm does not carry, and withdrawal is absent from all 158 samples, Q117's prediction, run: an MLL1 N-terminal ChIP and a FLAG ChIP of the tagged fusion in the same murine cells diverge 1.83-fold at Myb (0.6645 against 0.3623) and agree to within 4% at Meis1 (0.112 against 0.1171) - the direction is exactly as predicted and the pre-registered numeric threshold for Myb still fails, Q124's criterion H2 was written to be able to void Q124's own conclusion, and it cannot - because it is scored on a contrast in which the control genotype shows no effect to lose. Its raw arithmetic FIRES at 0.0813 against a 0.05 threshold, which reads as 'the knockout is a knockout' and is not evidence of anything, Re-score all nine GSE260742 same-condition floors under two further external stratifiers already binned on disk - KDM2B DMSO and the Oct-2024 input - to decide whether the twentyfold arm span is a property of the replicate pairs or of the stratum definition. Zero downloads, one re-run of an existing script., Re-score every committed within-sample share ratio at S_FLAG against Q117's eviction quartiles - the bins are already on disk in five scratch directories, no download and no wet work - and start with the H3K79me2 PCGF1 arm, because cycle 83's central null is the one most likely to change, RS4;11 - the human KMT2A::AFF1 index-fusion line - has a deposited H3K79me2 pair, a MENIN pair, an MLL1-N pair, a DOT1L pair and an input under the same menin inhibitor in GSE127507. Every track needed to run this beat's test on the index fusion exists. The download is 1.99 GB measured by HEAD request, and nobody has run it., SGC0946 removes 86% of H3K79me2's share at the fusion's own sites and the four eviction quartiles are 0.1396 / 0.1453 / 0.1452 / 0.1360 - span 0.0036, Spearman +0.400, a bootstrap interval spanning zero - so a very large effect on this mark can carry no eviction-depth structure whatever, Six GSE260742 samples titled 'NTvKO ... H2AK119ub rep1' carry ChIP_FLAG filenames AND a ChIP_FLAG string in the TDF header their own depositors' igvtools wrote - two of three independent sources say FLAG, the shape test is near-indecisive, and the channel of half the replicate arm cannot be settled from the deposit, Stratifying by eviction depth does NOT rescue H3K79me2 as a resistance-detecting readout: at the most-evicted quartile the PCGF1-knockout and the control are 0.6123 against 0.5860, a difference of 0.0263, while the whole genotype effect sits at the LEAST-evicted quartile (-0.1401) - the opposite of eviction-graded, Tested in an independent deposit with an epitope-tagged fusion: of the pivotal paper's three named losers under a menin inhibitor, MEIS1 and JMJD1C reproduce and MEF2C does not; of its three named retained loci, the HOXA cluster reproduces and MYB does not - and GAPDH, the negative control, loses half its share too, The cheap experiment nobody has run: an MLL1 C-TERMINAL ChIP beside the standard N-terminal one in a single human KMT2A-rearranged line under a menin inhibitor - the C-terminus is lost in the fusion, so the difference of the two tracks is the fusion, with no tag and no knock-in, The cheapest unrun test in this beat's queue: GSE260742 deposits an H3K79me2 ChIP under the DOT1L inhibitor SGC0946 in the same murine MLLAF9 cells as its FLAG and MLL1 tracks, which removes the catalysis directly instead of indirectly and would separate 'the mark is stable' from 'residual DOT1L is sufficient' - one download, The chromatin exchange at a menin-inhibited fusion site, measured in one experiment: H2AK119ub and H3K27me3 in, H3K4me3 out, in proportion to how much fusion left - and PCGF1 loss attenuates the in-fill without abolishing it, The deposit names its first replicate "paired" and its later ones rep1/rep2/rep3, so GSM8123403 "H2AK119ub paired" and GSM8123420 "H2AK119ub rep2" are one condition in two replicates - which this project has called "the two processing conventions" since cycle 88. GSE260742 holds 17 genuine same-condition replicate groups, 13 of them histone marks, refuting the claim repeated since cycle 100 that it has none and never will, The eviction quartiles are defined in the 'MLLAF9 Control' FLAG culture and the H3K79me2 they stratify is measured in the 'NTvKO Control' and 'NTvKO PCGF1 KO' cultures - zero shared libraries - so the commonest way a stratified analysis fools itself is structurally unavailable here, The experiment this beat could not do: an epitope-tagged KMT2A::AFF1 ChIP with a Drosophila spike-in, in a human infant-derived B-lineage line, at three replicates and two menin-inhibitor exposures - the design that would turn a within-sample ranking into an absolute occupancy measurement, The first replicated share ratio this corpus has ever had: two title-declared replicates of ONE murine culture span 0.2623 on the same quantity - 4.5x the 0.0584 between-culture spread this corpus uses as its indifference threshold - and they disagree in SIGN about whether the PCGF1 knockout blunts the DOT1L-inhibitor response, The first same-condition H3K79me2 noise floor measured anywhere in this thread: 11.17% at D1 and 26.44% at the TOP 0.1% stratum, against 45.72% and 62.63% for H2AK119ub in the same arm, the same condition and the same pairing shape - 4.1x tighter. A mark floor transferred from H2AK119ub does not describe H3K79me2., The four eviction quartiles are not equivalent stretches of genome: all three GSE260742 input tracks give a monotone increasing share across Q1 to Q4 at Spearman -1.000, 1.15 / 1.17 / 1.13-fold - small, real, in the same direction as the H3K79me2 gradient, and it cancels exactly in a within-genotype drug/vehicle ratio, The GSE260742 H2AK119ub pair that shares a processing state - rep2 against rep3, both bare `.tdf` - is not an unconfounded replicate pair: rep3 was submitted 2024-10-01 and rep2 2024-03-03, through a different file-naming pipeline. No pairing in this deposit isolates replicate variance alone., The H3K27me3 and H3K4me3 NTvKO arms - twelve files nobody had ever binned, listed as second priority - carry the largest effect in this cycle (H3K27me3 2.0007 at the most-evicted quartile, collapsing to 1.0134 in the PCGF1 knockout) and the free internal control that rules out a library artefact (H3K4me3 runs the opposite way, Spearman -1.000), The H3K79me2 eviction gradient against a DMSO-decile-matched null drawn from outside the fusion's sites at the same quartile sizes: the null is flat at 1.030-1.057, none of 20 draws reached the observed span of 0.6005, and the residual retains Spearman -1.000 - but the restricted-population trap is real and the baseline IS graded 2.2-fold, The H3K79me2 eviction gradient is not a ratio of sums dominated by a few bins: the MEDIAN per-bin log2 share ratio is itself monotone at Spearman -1.000, the fraction of bins individually below 1 falls 89.3% to 24.7%, the bootstrap span interval is [-0.651, -0.545], and no single bin moves the span by more than 0.0057, The menin inhibitor's effect on H3K79me2 is a REDISTRIBUTION within the fusion's own site set, not a loss from it: the share falls 41.4% at the quartile the fusion is most evicted from and RISES 18.7% at the quartile it is least evicted from, in murine MLL-AF9, The mid-occupancy floor was computed by three consecutive beats, stored in two committed results files for every contrast, and read as data by none of them - because it was introduced as a normaliser, so no criterion was ever written about it and no verdict ever mentioned it, The mouse fusion fraction is 0.73 and the human proxy fusion fraction is 0.21-0.24 - a three-fold gap in the direction a retroviral over-expression model predicts, so the one place the masking factor can be calibrated against a tag is the one place its value does not transfer, The only deposited H3K79me2 kinetic series for a menin inhibitor is non-monotone - a minimum at 16 h and a recovery at 24 h that survives common support and two rank statistics - and its 24 h point shows MORE redistribution than the 48 h main-arm sample, so the 48 h/72 h asymmetry in cycle 82's F stands and F keeps the words lower bound, The premise that GSE260742 'carries a ChIP-Rx spike-in' does not survive reading the deposit: zero of 158 sample records mention a spike-in, the only ChIP-Rx experiment in the paper is an H2AK119ub panel that is not identifiable among the deposited samples, and 152 of 158 samples ship only an unnormalised igvtools .tdf, The residual fraction this graph says is absent from the entire literature, measured: FLAG-MLLAF9 enrichment at its own top 1,228 sites falls from 15.9x the genome median to 3.0x under VTP50469 - about 18% residual - and 12.8% of those sites stay in the top 0.1%, The same-condition H2AK119ub noise floor in GSE260742 spans twentyfold across the deposit's three experimental arms - D1 |R-1| of 1.89% in the PCGF1-KO clone, 10.49% in the non-targeting control and 37.04% in the MLLAF9 Control arm - at one mark, one lab, one pairing shape and one external stratifier. A single same-condition floor for this deposit does not exist., The shared-epitope masking calibration, run against a FLAG tag and then refuted by its own control: the MLL1 antibody reports 2.23-fold more retained signal than the tag at the fusion's own sites, but MLL1-dominated non-fusion sites fall HARDER than the fusion does (0.1069 against 0.1793), so no mixture of fusion and wild-type MLL1 can produce the observed 0.3995, The universe-overlap gate this thread inherited from cycle 107 divides by a track and multiplies by the stratifier, so it also measures whether both libraries are as deep as the stratifier - a property of the stratifier, not of the pair. It aborted 13 of 16 comparisons here, and cycle 107's own pair passed it by 0.25 percentage points while losing more bins to the stratifier than any pair it refused., This beat pre-registered 'the observed span exceeds the matched-null span by >= 5x' and never said WHICH summary of a 20-draw null: against the maximum it is 3.71x and the criterion fails, against the median it is 30.0x and it passes. The conservative reading was scored and the defect is recorded rather than repaired., This cycle's word 'silenced' means two repressive marks up and one active mark down at 2 kb resolution - GSE260742 deposits no RNA-seq, no nascent transcription and no reporter for either NTvKO genotype, so whether the vacated sites stop being transcribed is unmeasured in the experiment that established they are marked for silencing, Two deposit defects in one beat, both invisible to the check in use: a genome-build verdict read from a file's declared extents names its CONTAINER, and the furthest coordinate carrying signal names its DATA - and a clause that cannot return the other answer has been reported as a passing check in four beats, Two drugs, two spatial mechanisms on one mark: catalytic DOT1L inhibition de-concentrates H3K79me2 out of its peaks and into mid-occupancy chromatin, while menin-interface inhibition lowers the mark's share at the peaks AND at mid-occupancy, so its effect is relocation rather than a smaller version of the same change, Two processing batches inside one GEO series: the FLAG and KDM2B tracks put 1.7-1.9% of their coverage in ENCODE blacklist regions and the BCOR, RING1B and input tracks put 14-46%, with a single 2 kb bin at chr2:98,666,000 carrying 18.1% of the BCOR SGC0946 track on its own, Verified against the live records rather than inherited: HUDEP-2 and mESC each carry a DMSO and an SGC0946 H3K79me2 sample and no VTP-50469 sample, so K562 remains the only fusion-free line in GSE260742 able to score a menin-inhibitor effect - and cycle 82's request for a second and third fusion-free line cannot be met from this series, What fills the space: RING1B and BCOR arrive at the fusion's sites under BOTH drugs, but the DOT1L inhibitor puts them everywhere on the fusion's territory while the menin inhibitor puts them only in the quartile where the fusion is most completely evicted - and at Meis1 RING1B rises 16-fold, When a deposit has no recoverable spike-in, two comparisons still work and one does not: a within-sample rank and a within-sample share both cancel the unknown scale factor, a fold change does not - and the share is the one that also has no ceiling, With H3K79me2 added, the eviction-depth panel over one set of 1,228 bins under one drug is H3K27me3 +1.000, H2AK119ub +1.000, H3K4me3 -1.000 and H3K79me2 -1.000 - a clean repressive-up / active-down split that no shared library, blacklist or regression artefact can produce, Within one arm, swapping the processing-marker confound for a submission-batch confound takes the same-condition H2AK119ub floor from 37.04% to 16.84% at D1 (tighter, 0.45x) and from 37.37% to 61.89% at the TOP 0.1% stratum (worse, 1.66x). The brief's question - tighter, wider, or unchanged - has no single answer, and it is worse exactly where this thread scores its claims.
42. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    Claim-support check: IDF-weighted overlap between every claim and the abstract it cites

    · Oncofusion knowledge graph (internal tooling) 2026
    used by A checker for the error the citation checker cannot see
43. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    Europe PMC RESTful article search, 25 BODY: intersections run and title-screened 2026-09-22 (scratchpad q115_proceedings.json)

    · Europe PMC 2026
    used by A congress abstract book is one indexed document containing thousands of unrelated abstracts, so it matches almost any conjunction: across 18 Europe PMC full-text intersections read to completion, 83 of 321 records (25.9%) are proceedings, ranging 0% to 50% per query, Re-bound at full text: no study of MLLT1/ENL function in cerebellum or developing brain exists, and 10 of the 22 full-text hits are congress abstract books - the weakest-bounded absence in the corpus is now the best-bounded of the four ENL ones
44. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    Europe PMC RESTful search, paged retrieval of BODY:(LEDGF aliases) AND BODY:(20 menin-inhibitor names), pages 1-3, 2026-09-22

    · Europe PMC 2026
    used by A congress abstract book is one indexed document containing thousands of unrelated abstracts, so it matches almost any conjunction: across 18 Europe PMC full-text intersections read to completion, 83 of 321 records (25.9%) are proceedings, ranging 0% to 50% per query, Re-bound at full text: the LEDGF-after-menin-failure intersection goes from 0 records at abstract level to 116 at full text and 90 with resistance terms, and the 50 read are review articles - no LEDGF-directed agent has been tested in a menin-inhibitor-resistant model
45. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    PubMed E-utilities esearch counts for SET-NUP214 against histone-mark vocabularies, query translations read back

    · NCBI PubMed 2026
    used by A counted PubMed absence is bounded by ABSTRACTS, not by papers - and the measurement you are looking for is usually a supporting figure, SET::NUP214 recruits DOT1L by a different surface from CALM-AF10 and KMT2A::MLLT10: the founding paper tested for AF10 in the complex and did not find it
46. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    Opposed-claims detection: can lexical overlap find a contradiction between two papers that share no identifier?

    · Oncofusion knowledge graph (internal analysis) 2026
    used by A detector for cross-paper contradictions, and the measurement that says it does not work
47. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
    AUGMENT-101: A Study of Revumenib in R/R Leukemias Including Those With an MLL/KMT2A Gene Rearrangement or NPM1 Mutation

    Syndax Pharmaceuticals · ClinicalTrials.gov 2026
    used by A documented absence: not one endocrine outcome measure across the three lead menin-inhibitor trials, AUGMENT-101 (NCT04065399) — the registration trial, and what 22.8% actually means, AUGMENT-101: the partner composition of the trial that approved a menin inhibitor, with KMT2A::AFF1 at n=2 evaluable, Bounded absence: not one menin-inhibitor study, trial or preclinical paper has ever recorded reciprocal fusion status, Breakthrough therapy and accelerated approval: the menin field took the fast review and declined the shortcut, CALM-AF10 was meant to separate the menin axis from the HOXA programme - and the indirect evidence has now turned against separability, because KMT2A-free HOXA lesions keep proving menin-DEPENDENT, Counted absence: no menin-inhibitor trial has ever reported a response RATE in NUP98-rearranged patients, Menin inhibitors are prescribed on 'KMT2A-rearranged', a category that contains proteins as different as KMT2A::AFF1 and KMT2A::USP2, Model: predicting who responds to a menin inhibitor - no patient-level data is public, so this cannot be trained either, No menin-inhibitor or DOT1L-inhibitor trial has ever named this fusion - and the DOT1L trials excluded it by construction, PICALM::MLLT10 (CALM-AF10), t(10;11)(p13;q14) - the DOT1L-dependent HOXA fusion that contains no KMT2A, Syndax Pharmaceuticals — revumenib (SNDX-5613), the only menin inhibitor licensed down to infancy, Systematic enumeration of menin-inhibitor resistance mechanisms: seven entries written, thirteen candidates rejected with reasons, The counter-argument: the menin field is an adult field, and the infant studies were withdrawn, The rare paediatric disease priority review voucher: a $150-200m bearer bond, lapsed for 14 months, now live to 2029
48. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
    KOMET-001: First in Human Study of Ziftomenib in Relapsed or Refractory Acute Myeloid Leukemia

    Kura Oncology, Inc. · ClinicalTrials.gov 2026
    used by A documented absence: not one endocrine outcome measure across the three lead menin-inhibitor trials, Breakthrough therapy and accelerated approval: the menin field took the fast review and declined the shortcut, Kura Oncology — ziftomenib (KO-539): an adults-only sponsor that quietly supplies infant trials, No menin-inhibitor or DOT1L-inhibitor trial has ever named this fusion - and the DOT1L trials excluded it by construction, The counter-argument: the menin field is an adult field, and the infant studies were withdrawn
49. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
    cAMeLot-1: A Phase 1/2 Study of Bleximenib in Participants With Acute Leukemia

    Janssen Research & Development, LLC · ClinicalTrials.gov 2026
    used by A documented absence: not one endocrine outcome measure across the three lead menin-inhibitor trials, Bleximenib (JNJ-75276617) — menin inhibitor with a cleaner cardiac profile, and the only one with a 2-year-old floor, Counted absence: no menin-inhibitor trial has ever reported a response RATE in NUP98-rearranged patients, Johnson & Johnson / Janssen — bleximenib (JNJ-75276617): paediatric from age 2, but not infants, and the dedicated child study was withdrawn, No menin-inhibitor or DOT1L-inhibitor trial has ever named this fusion - and the DOT1L trials excluded it by construction, The cheapest study that would settle this: nine assays and a hand radiograph, on children already being followed
50. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
    NCBI GEO GSE260742, GSM8123359 / GSM8123360 / GSM8123361 (MLLAF9 FLAG ChIP-seq, DMSO / SGC0946 5 uM / VTP50469 500 nM), re-binned to 2 kb from the deposited igvtools .tdf

    · NCBI Gene Expression Omnibus 2026
    used by A DOT1L inhibitor, which does not touch the menin-KMT2A interface, removes 70% of the fusion's enrichment from its own strongest sites - two thirds of what a menin inhibitor removes - so fusion chromatin residency is partly downstream of H3K79 methylation rather than only of menin
51. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    NCBI GEO GSE260742 locus table over UCSC mm10 refGene windows, analyses/q117-gained-fusion-occupancy/locus-table.json

    · NCBI Gene Expression Omnibus 2026
    used by A DOT1L inhibitor, which does not touch the menin-KMT2A interface, removes 70% of the fusion's enrichment from its own strongest sites - two thirds of what a menin inhibitor removes - so fusion chromatin residency is partly downstream of H3K79 methylation rather than only of menin, Rule 62 applies to a genome browser: the UCSC search API returns several tracks and many rows per gene, 'take the longest' put Hoxa9 and Meis1 on chr1, only the refGene track carries the SYMBOL in posName, and xenoRefGene returns a human-aligned Six1 on mouse chr17
52. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    REVUFORJ (revumenib) tablets, for oral use - US Prescribing Information (SPL version 5, published 20 Feb 2026)

    Syndax Pharmaceuticals, Inc. · DailyMed / FDA Structured Product Label, setid 6eb3cdbc-0e74-477d-82d6-3bb172d3f63f 2026
    used by A drug that fixes diabetes in mice, and neither label contains the word glucose, Four different ways to lose menin, and the graph has been treating them as one, Hypothesis: the same drug is protective in the islet and harmful in the parathyroid and the growth plate, Menin builds bone early and blocks it late: two opposite jobs in the same lineage, Menin is required in neural crest for the palate to close - and both menin inhibitors cause cleft palate in animals, Menin is the switch that lets TGF-beta turn parathyroid hormone down - and it works on secretion, not just growth, Nobody needs to enrol anyone: the PTH question is a regression on data that already exists, The cheapest missing experiment in this beat: put revumenib on a parathyroid cell, The label says monitor bone growth in children. Not one paper has asked what menin does in a growth plate, The parathyroid does not need the rest of the body: deleting menin only there causes neoplasia and hypercalcaemia, Three extra tubes: the glucose study that answers the inversion in children, Twelve weeks of a menin inhibitor in diabetic mice: more beta cells, lower fasting glucose, and nobody looked for a tumour, What the labels measure in a treated child: bone growth, in one sentence, and nothing else endocrine
53. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.database
    analyses/q204-failure-mode-atlas/atlas.json, atlas_extra.json, atlas_na.json validated by atlas.py on 2026-10-14 (corpus of 3,128 nodes)

    · this repository 2026
    used by A failure-mode atlas for drugging the KMT2A::AFF1 axis: 90 modes over a 10 x 11 grid of layers and mechanisms, 46 cells argued not applicable, and eight modes the corpus has no entry for
54. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    analyses/menin-under-revumenib-in-an-infant-patient-pair - pre-registered re-analysis of GSE252378 samples 148752 and 152985

    Q143-menin-under-drug-in-infants (Oncofusion Knowledge Graph) · this repository 2026
    used by A filter-free promoter ranking with no peak caller and no target list recovers 40 of the top 50 published oncoprotein-target genes in the drug-free infant sample, and 0 of 50 in the drug-exposed one, An occupancy readout returned background in every channel of the only drug-exposed patient sample it has been tried in - so a negative from such an assay cannot currently be told apart from an assay failure, ENL redistributes rather than persists: it is the only channel that holds more of its signal at the myeloid-program peaks than at the B-lineage ones, by 2.06x in my instrument and 1.26x in the authors' own - and in absolute terms it falls further than menin does, In a patient, the menin-fusion association is demonstrable before the drug and unmeasurable after it - and the asymmetry is what this pair establishes, not a rate of menin loss, In drug-naive primary infant KMT2A::MLLT1 B-ALL, menin is as enriched as the fusion N-terminus at the top fusion promoters (2.73x vs 3.12x over an accessibility reference) while the wild-type C-terminal reporter sits at background (1.10x), In the revumenib-exposed relapse sample nothing is enriched above a non-specific reference - not menin, not the fusion N-terminus, not ENL - and the sample's own top-ranked promoters are mitochondrial pseudogenes, Menin does not outlast the fusion under revumenib in primary infant material - 0.586 against 0.485 at oncoprotein peaks, a ratio of 1.21 against a pre-registered threshold of 1.5 - so the fusion-independent menin plateau seen after a degron does not reproduce here, Menin, ENL and DOT1L are CUT&Tag and the four KMT2A antibodies are CUT&RUN, run three weeks apart - so every menin-versus-fusion comparison in this pair crosses both assay and batch, Neither infant sample has an IgG control, and neither of their two sequencing batches contains one - so every number from this pair is confounded with chromatin accessibility, Spike-in-normalised menin CUT&RUN with paired N- and C-terminal KMT2A antibodies in an infant KMT2A::AFF1 sample, on and off drug, in the same disease state, The KMT2A N-terminal antibody reports wild-type plus fusion and the C-terminal antibody reports wild-type alone - which makes an N-versus-C difference a fusion-specific readout with no tag and no engineering, The menin that matters in an infant KMT2A-rearranged leukaemia is the menin on the fusion, and the N-minus-C antibody difference can measure it in any patient sample without a tag, The menin, ENL and DOT1L replicates of the drug-exposed relapse sample agree with each other at Spearman 0.11 to 0.25, below the floor this project set in the previous cycle, so no quantitative claim from that arm survives
55. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paperPMID 42389523
    Case Report: Safety and efficacy of blinatumomab in combination with donor lymphocyte infusion for the prophylaxis of relapse following allogeneic hematopoietic stem cell transplantation in a pediatric patient with acute lymphoblastic leukemia.

    Li YT, Li Y, Wang LN, Huang LB · Frontiers in Immunology 2026
    used by A five-year-old with KMT2A::USP2 ALL whose MRD would not clear until blinatumomab, then transplant, then more blinatumomab, KMT2A::USP2, inv(11)(q23.3q23.3) - the minor-BCR fusion that keeps the PHD fingers
56. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
    analyses/the-mll1-channel-two-by-two/run.py and results.json, this repository, computed 2026-09-23 from GEO GSE173599 MACS peak tables (GSM5271717-720 and GSM5271712-715); deposit publication PMID 35245435 verified as Olsen, Godfrey et al., Molecular Cell 2022, PMC9044330

    Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
    used by A full 2x2 on the MLL1 channel in one run - DMSO, EPZ5676 1 uM, VTP-50469 100 nM and both - closes cycle 83's thirty-cycle-old threat: at the fusion's strongest MLL1 sites the menin inhibitor removes 58.3% against the DOT1L inhibitor's 17.5%, a 3.33-fold separation at TEN TIMES the DOT1L dose Q136 could test, with sequencing depth running opposite to the effect in every arm
57. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Ensembl REST API /sequence/region/human/11:118336456..118626832, coord_system_version=GRCh38, content-type text/plain

    Ensembl · Ensembl REST 2026
    used by A genome-wide DAMAGE-STATE occupancy dataset for CtIP and MRE11 does exist and is public - GSE218590, 58 samples, +/- 4-OHT and a light-activated Cas9 time course - and it cannot answer the KMT2A question, because there is not one AsiSI site in the 90,377 bp of the gene
58. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    analyses/q37-archive-sweep-at-the-kmt2a-locus - results.json, tracks SRX20732396-406 (RBBP8, U2OS) and SRX18353285-290 (MRE11A, HEK293)

    Q37-archive-sweep · this graph 2026
    used by A genome-wide DAMAGE-STATE occupancy dataset for CtIP and MRE11 does exist and is public - GSE218590, 58 samples, +/- 4-OHT and a light-activated Cas9 time course - and it cannot answer the KMT2A question, because there is not one AsiSI site in the 90,377 bp of the gene, ENCODE spells zero as HTTP 404, ReMap spells it as HTTP 401, and GEO spells it as a well-formed count of zero under a retired symbol - three archives, three spellings, and two of them are indistinguishable from a broken connection, Outside ENCODE the resection layer is not one absence but four: 31 BRCA1, 45 CBX3, 22 KDM5A, 14 MRE11 and 8 CtIP ChIP-seq experiments exist in ChIP-Atlas, DNA2 and RAD50 have one each, and only BARD1 and EXO1 have none in any of the three archives
59. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
    Q154 streaming gene-body extractor and instrument gate

    Iqbal Lab oncofusion knowledge graph, beat Q154 · this project 2026
    used by A genome-wide gene-body occupancy matrix over 133 libraries costs one fragment BED of peak disk, not 133 - stream, bin in flight, delete
60. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
    ENCODE Portal: all 13 released human BRCA1, CBX3 and CBX1 TF ChIP-seq experiments, signal and peak files, downloaded and analysed 2026-09-22

    ENCODE Project Consortium · ENCODE Portal 2026
    used by A genome-wide null flatters any interval that belongs to an enriched class: the KMT2A promoter ranked in the top 0.07% of the genome and is 1.24x an ordinary housekeeping promoter, and the exon-11 element ranked in the top 3% and sits BELOW the median CTCF anchor, What is left of the resection absence after the compute was done: the resting state is now MEASURED and negative in 13 ENCODE datasets, so the whole question collapses onto the DAMAGE state, which is never-measured for all six proteins in every cell type on Earth
61. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Antibody Registry record 3713372, TEL-AML1 (ETV6-RUNX1) Antibody, clone TEL_6F2 (RRID:AB_3713372)

    · SciCrunch Antibody Registry 2026
    used by A junction-specific monoclonal antibody against a paediatric leukaemia fusion oncoprotein exists: clone TEL 6F2, raised against a 12-mer spanning the ETV6::RUNX1 fusion site and used for genome-wide ChIP, Raise a KMT2A::AFF1 junction-specific monoclonal by the published TEL 6F2 protocol - rat immunisation with a junction peptide, ELISA and western screen, then a ChIP test against the wild-type-parent controls that the ETV6::RUNX1 campaign never ran
62. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Antibody Registry full-text search, q=SS18-SSX (returns RRID:AB_3665243, SS18-SSX (E9X9V) XP Rabbit mAb, Cell Signaling Technology)

    · SciCrunch Antibody Registry 2026
    used by A junction-specific monoclonal antibody against a paediatric leukaemia fusion oncoprotein exists: clone TEL 6F2, raised against a 12-mer spanning the ETV6::RUNX1 fusion site and used for genome-wide ChIP
63. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Antibody Registry record 490962, Bcr-Abl (b2a2 Junction Specific) (L99H4) Mouse mAb (RRID:AB_490962)

    · SciCrunch Antibody Registry 2026
    used by A junction-specific monoclonal antibody against a paediatric leukaemia fusion oncoprotein exists: clone TEL 6F2, raised against a 12-mer spanning the ETV6::RUNX1 fusion site and used for genome-wide ChIP, Re-bound in the matching instrument: the Antibody Registry holds no junction-specific antibody for any KMT2A fusion under five spellings, while the same query template returns the synovial sarcoma clone as a positive control and turns up two leukaemia-fusion junction reagents nobody here had found
64. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42005175
    Sequential Transformation of Polycythemia Vera to Myelofibrosis and KMT2A-Rearranged Acute Myeloid Leukemia Treated With Revumenib: A Rare Case of Clonal Evolution

    Ahmad CM, Kastle RA, Zargarian L, Sawalha Y, Dalia S · Cureus 2026
    used by A KMT2A fusion acquired at seventy-three: KMT2A::ELL AML arising out of polycythaemia vera, and a menin inhibitor used to hold the line
65. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    IntAct PSICQUIC MITAB 2.7 export for id:Q8TEK3 (DOT1L), 49 rows, fetched 2026-09-21, read alongside Hein et al. 2015 supplementary Table S2 prey.IDs column

    · EMBL-EBI IntAct 2026
    used by A mass-spectrometry protein group's leading identifier is not the interactor, and curating it puts dubious isoforms and unreviewed accessions into the interaction databases, IBTK - the only one of the six DOT1L spokes with a reciprocal test, and its own pulldown in the same study does not find DOT1L, MAPK7 (ERK5) - a DOT1L partner on one 1:365 spoke of one atlas pulldown, and a kinase with real leukaemia literature that has never been connected to DOT1L, NIN (ninein) - a centrosomal scaffold recorded as a DOT1L partner under a UniProt isoform that UniProt itself calls dubious, PARP16 - graded A- in the DOT1L pulldown, the best of the six, and a tail-anchored ER membrane protein that cannot be in the nucleoplasm, The six unwritten DOT1L partners are not six findings - they are nine spokes of ONE GFP-DOT1L pulldown, and the source paper already graded them
66. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB Q9P2D0 (reviewed, 1353 aa) and E7EPI0 (unreviewed, 1338 aa), both human IBTK

    · UniProt Consortium 2026
    used by A mass-spectrometry protein group's leading identifier is not the interactor, and curating it puts dubious isoforms and unreviewed accessions into the interaction databases, IBTK - the only one of the six DOT1L spokes with a reciprocal test, and its own pulldown in the same study does not find DOT1L
67. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
    Is '100% of the directed-evidence interactome is written' still true? No. It is 68.5%.

    · Oncofusion knowledge graph (internal analysis) 2026
    used by A mass-spectrometry protein group's leading identifier is not the interactor, and curating it puts dubious isoforms and unreviewed accessions into the interaction databases, Atlas-only interactome partners are predicted by how abundant the prey is, not by which bait was pulled - and if so this graph's interactome denominator is inflated
68. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 42578372
    The tumor suppressor menin is a key scaffold mediator for homologous recombination repair.

    Lian A, Zhu X, Jin B, Gu G, Xu B, Zhang L, Lin X, Zhu X, Jin G · Nucleic Acids Res 2026
    used by A menin degrader removes six surfaces; a pocket inhibitor blocks one and touches a second - and in a child receiving TOP2 poisons that difference is a repair-competence difference nobody has measured, A residue axis for menin, derived from coordinates: three spatial lobes, not four named ones, and the sharpest seam in the protein runs through the drug pocket, ADJUDICATION: the SvID and the gamma-H2A.X motif are OVERLAPPING BUT NOT THE SAME SURFACE - and the premise that made the question urgent, that the SvID is distinct from the MLL cavity, is FALSE at the SvID's own boundary, Does a menin inhibitor break homologous recombination? Test the menin - SETD2 - LEDGF - CtIP chain in a blast, before the next genotoxic combination is designed, EXO1 (Q9UQ84, 846 aa) - the long-range resection nuclease at the bottom of the CTCF cascade, and the point where this graph's two independent 2026 HR papers meet: CtIP starts the resection, EXO1/DNA2 finish it, Menin stimulates homology-directed repair through Chk1 and the region required is 1-228: a 2010 paper that is cited in no entry of this graph and in no reference list of the 2026 paper that reports the same conclusion, Menin-SETD2, mapped to the NTD and Palm and tooled with a peptide at menin 231-248: the third partner assigned to the same half of menin, and the writer arm of the HR scaffold, Occupancy-matched three-arm comparison of ziftomenib, revumenib and bleximenib on gamma-H2A.X resolution and HR efficiency in a KMT2A-rearranged B-lineage line, with the peptide competition as the biochemical pre-test, RBBP8 / CtIP - the only unwritten core-complex partner with a DIRECT-interaction record, and the graph already held its mechanism without holding the protein, RPA1 (RPA70) - the large subunit of replication protein A, which carries most of the ssDNA-binding folds and the N-terminal fold that hands the lesion to ATR, The menin acidic shelf, Glu366-Asp370-Val371: the counter-charge that MBM2 binds, resolved in 3U88 after all - and the one patch where the drug pocket, the SvID and the gamma-H2A.X motif all touch, The menin arm and the CTCF arm are not two stories: menin/SETD2/H3K36me3/LEDGF licenses CtIP short-range resection and CTCF/H3K9me/HP1-gamma licenses BRCA1/BARD1-EXO1/DNA2 long-range resection, which are the two consecutive halves of ONE reaction - so a menin inhibitor and an H3K9-axis inhibitor are acting on the same pathway at different steps, The menin-gamma-H2A.X motif, menin Val371-Gln400: necessary, sufficient as a peptide, and the nine residues at its centre have never been seen in any of the 69 menin structures, The menin-inhibitor class splits at Val371: three of four clinical compounds contact the first residue of the gamma-H2A.X motif and bleximenib does not, so DNA-repair effects should separate from KMT2A-displacement effects within the class, The two co-IPs that settle whether menin's SUV39H1 window and its gamma-H2A.X motif are one surface: swap each laboratory's construct into the other's assay, UniProt's menin 214-390 FANCD2 region is not a surface: it spans three spatial lobes and 29.9% of menin's solvent-accessible area, so its twenty-residue overlap with the gamma-H2A.X motif is what chance gives you
69. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    analyses/q34-menin-surface-geometry/run.py and results.json - menin surface geometry over all 69 O00255 PDB deposits

    Q34-menin-surface-map · Oncofusion Knowledge Graph 2026
    used by A menin degrader removes six surfaces; a pocket inhibitor blocks one and touches a second - and in a child receiving TOP2 poisons that difference is a repair-competence difference nobody has measured, ADJUDICATION: the SvID and the gamma-H2A.X motif are OVERLAPPING BUT NOT THE SAME SURFACE - and the premise that made the question urgent, that the SvID is distinct from the MLL cavity, is FALSE at the SvID's own boundary, The menin-gamma-H2A.X motif, menin Val371-Gln400: necessary, sufficient as a peptide, and the nine residues at its centre have never been seen in any of the 69 menin structures
70. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    PDB 8VA6: Menin in complex with Ziftomenib (KO-539)

    RCSB Protein Data Bank depositors · RCSB Protein Data Bank 2026
    used by A menin degrader removes six surfaces; a pocket inhibitor blocks one and touches a second - and in a child receiving TOP2 poisons that difference is a repair-competence difference nobody has measured, The menin small-molecule pocket as a coordinated surface: four clinical-stage compounds, 21-26 contact residues each, and bleximenib sitting on a different part of it from the other three
71. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
    PubMed esearch: (revumenib OR ziftomenib OR bleximenib OR "menin inhibitor" OR "menin inhibitors" OR "menin inhibition" OR SNDX-5613 OR KO-539 OR VTP-50469 OR MI-503)[tiab] AND (lipid OR lipids OR cholesterol OR triglyceride OR triglycerides OR steatosis OR lipogenesis OR LXR)[tiab]; run 2026-09-21

    · NCBI PubMed 2026
    used by A menin inhibitor is a partial LXR agonist, and no trial in the class has looked
72. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    A reusable two-null selectivity test for any gene set, run on FACT and the histone chaperone layer

    · Oncofusion knowledge graph (internal analysis) 2026
    used by A null that costs one command, run first rather than afterwards - and SSRP1 is genuinely LESS required in KMT2A-rearranged lines, while the rest of the chaperone layer is not much of anything
73. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
    analyses/q116-h3k9-axis-at-the-locus/results_scored.json, controls C5 and C5b; pre-registered in PREREGISTRATION.md section 5 before any of it was measured

    Q116-h3k9-axis-at-the-locus · oncofusion-knowledge-graph 2026
    used by A panel of many antigens in one cell type is the input track the deposit did not provide: eight H3K9-axis proteins in MCF-7 all read the whole KMT2A locus elevated (six-non-anchor medians 49.8-72.4) and all read the KMT2A promoter at the 87.6th-99.9th percentile - so the shared component has to be subtracted before any single antigen's number means anything, NEITHER, and it is the closest call this project has had: KAP1/TRIM28 reads the 90.62nd, 81.85th and 89.45th percentile of MCF-7 CTCF anchors at the exon-11 element in three of four tracks - satisfying the first clause of the pre-registered PRESENT rule - and the six-non-anchor-interval clause blocks it at 65.87, The H3K9 machinery's address at KMT2A is the PROMOTER, not the exon-11 breakpoint anchor: G9a, HP1a and KAP1 all read the top decile of active promoters there in MCF-7 and KAP1's peak replicates a K562 peak to 267 bp, while five proteins across five cell types are ORDINARY or absent at the anchor - so every H3K9-directed intervention this graph has discussed is aimed 52 kb from where the machinery is, THE POSITIVE RESULT: the writer G9a, the reader HP1a and the scaffold KAP1 all sit on the KMT2A PROMOTER in MCF-7 at the 87.6th-98.9th percentile of 2,000 ACTIVE promoters - surviving the control that says an active promoter is ChIP-hot for everything - while none of the three is above the 75th percentile at the exon-11 anchor 52 kb away
74. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    ana-q116-g9a-sits-on-the-kmt2a-promoter-in-mcf7-and-is-ordinary-at-the-exon-11-anchor and ana-q116-trim28-comes-closer-to-firing-at-the-exon-11-anchor-than-anything-yet-and-the-specificity-clause-blocks-it

    Q116-h3k9-axis-at-the-locus · oncofusion-knowledge-graph 2026
    used by A panel of many antigens in one cell type is the input track the deposit did not provide: eight H3K9-axis proteins in MCF-7 all read the whole KMT2A locus elevated (six-non-anchor medians 49.8-72.4) and all read the KMT2A promoter at the 87.6th-99.9th percentile - so the shared component has to be subtracted before any single antigen's number means anything, AP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology, Blot all three PBX proteins in KMT2A::AFF1 B-lineage cells, then delete them singly and in combination - the family has never been tested together in any human leukaemia line, Does menin hold FANCD2 in a KMT2A-rearranged cell, and does revumenib let go? Nobody has asked — zero records over every menin-inhibitor name, GOLGA2 (GM130) - a Golgi matrix protein with 249 partners in the one yeast screen that reported it binding AFF4, which had five, Measured genome-wide, not just at the locus: the median WITHIN-CONDITION replicate disagreement between two ChIP-seq tracks at a 146 bp window is 23.48 percentile points across 56,171 CTCF anchors and 12 replicate pairs, rising to 57-83 points at the 90th percentile of anchors - and the KMT2A exon-11 element sits at the 35.8th percentile of that distribution, so it is NOT unusually noisy, 146 bp windows are, MTUS2 (TIP150) - a microtubule plus-end tracking protein with 153 partners in the one yeast screen that reported it binding AFF4, Of the 58 two-patient AFF1-exclusive genes, four - not one - are called exclusive by both instruments, and the surviving IRX1 claim does not beat a patient-level null (p=0.092) [CORRECTED by q171], Reading the same-source counter-evidence: 77 notes adjudicated, and the defect is not the one the metric measures, RUN: endogenous HP1a/CBX5 is DEPLETED at the exon-11 anchor (median 23.94th percentile) and has the lowest KMT2A gene-body reading of any protein in the panel (30.16) - while the phosphomimetic Bio-ChIP mutant 4E reads higher than the phospho-compromised 4A at all seven intervals AND that difference falls inside the genome-wide interquartile range of the same comparison, so it carries no information, RUN: KAP1/TRIM28 has a called peak overlapping the KMT2A promoter in MCF-7 in both hormone conditions, 267 bp of it inside the K562 TRIM28 peak Q110 found - and its nearest chr11 peak to the exon-11 CTCF summit is 52,230 bp away, RUN: KDM4B and KDM4C both read the 76.6th-79.7th percentile of MCF-7 CTCF anchors over the whole 90 kb KMT2A gene body - above the 25-track panel median of 61.39, LOCUS-WIDE ELEVATED - and both are ORDINARY at the exon-11 anchor with replicate spreads of 92.3 and 61.5 percentile points, THE DEFLATING NULL, RUN: matching on gene length AND MCF-7 expression roughly HALVES the H3K9me3 depletion over the KMT2A gene body - from the 7.6th-15.0th percentile of class-matched CTCF anchors to the 18.2nd-23.4th of 385 matched transcribed gene bodies - so the methyl-poor breakpoint domain is partly a fact about transcription and does not dissolve into one, The experiment this beat could not do: G9a, HP1a, KAP1 and H3K9me3 at the KMT2A promoter and exon-11 anchor in a KMT2A-rearranged B-lineage cell and a CD34+ progenitor, WITH an input arm, an IgG arm and a same-cell CTCF arm, The H3K9 machinery's address at KMT2A is the PROMOTER, not the exon-11 breakpoint anchor: G9a, HP1a and KAP1 all read the top decile of active promoters there in MCF-7 and KAP1's peak replicates a K562 peak to 267 bp, while five proteins across five cell types are ORDINARY or absent at the anchor - so every H3K9-directed intervention this graph has discussed is aimed 52 kb from where the machinery is, The H3K9me3 this graph argues about at the KMT2A exon-11 anchor is not deposited by SUV39H1: in SU-DIPG13 the mark sits at the 57.8th percentile of class-matched CTCF anchors while the writer reads an EXPLICIT ZERO in the same nuclei, and genome-wide the two barely track each other (Spearman 0.148, 0.078), The PBX requirement in KMT2A-rearranged leukaemia is a family property, and every experiment that shaped this graph's view of it removed one paralogue at a time, TRAF2 - a cytoplasmic NF-kB brake reported as an AFF4 two-hybrid hit, 97 partners in the screen that found it, and the live instance of the brief's own accession trap, UTP18 is not a KMT2A partner - it is WDR5, mis-curated into IntAct from the founding paper, and it manufactured the only KMT2A-bait gap in the interactome recount, Zero of eight unwritten AFF4/MEN1 partners has an experimentally coded nuclear localisation, and the two that are called nuclear are called it by inference
75. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 41730833
    Marrow leptin-LEPR signaling rewires mitochondrial oxidative metabolism to confer chemoresistance in acute myeloid leukemia.

    Liao X, Dai W, Xu X, Cai D, Tan M, Wang Z, Huang Y, Hou D, Liu J, Wang L, Wang J, Wang X, Zhang S, Lin X, Huang H · Cell Death Dis 2026
    used by A partial resolution: the leptin-LEPR chemoresistance model is murine MLL-AF9, but the two AML arms were never separated and the proteomics deposition is unreachable, Bone-marrow adipocyte — the fat cell that suppresses normal blood formation and feeds leukaemia a chemoresistance signal, LEPR - the marker that defines the SCF-supplying perivascular niche cell, the receptor for an adipocyte signal that blunts cytarabine, and a gene_exact query that returns the wrong protein first, Niche-mediated chemoprotection in KMT2A-rearranged leukaemia — real for cytarabine, unsupported for menin inhibitors, and the one hypoxia experiment points the other way
76. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    PRIDE Archive REST API project record request for PXD066693, and NCBI elink pubmed to gds, sra, biosample, bioproject for PMID 41730833

    · EMBL-EBI PRIDE Archive 2026
    used by A partial resolution: the leptin-LEPR chemoresistance model is murine MLL-AF9, but the two AML arms were never separated and the proteomics deposition is unreachable
77. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    analyses/menin-inhibitor-against-dot1l-inhibitors-in-one-deposit (PREREGISTRATION.md, run.py, results.json) - this repository, re-analysis of the ten deposited GSE149183 MACS2 peak tables

    · Oncofusion knowledge graph 2026
    used by A per-sample parameter nobody reads: MACS's fragment-size estimate varies 822-1112 bp across ten ChIPs from one sequencing run, and within the DOT1L arm peak count correlates +0.7 with fragment size and -0.8 with library depth - so the obvious confound is not the operating one, and the non-obvious one is, A tempting second finding that does NOT survive its control: about half of each treated sample's peaks are absent from DMSO, but the two independent VTP-50469 libraries share only 10.5% of their gained sites - against 7.7%, 8.1% and 10.6% for the DOT1L-inhibitor arms - so the gained peaks are threshold noise and there is no evidence of redistribution, CORRECTS A DISPATCH PREMISE: GSE149183's menin-versus-DOT1L-inhibitor head-to-head exists in MOLM13 (KMT2A::MLLT3) ONLY. RS4;11 - the index fusion - is in the series but carries H3K79me2 alone, with no VTP-50469 arm, no MEN1 ChIP and no DOT1L ChIP, and the two arms are on different genome builds, First deposit to put both drug classes on the MEN1 channel in one experiment: VTP-50469 leaves ZERO of the 19 DMSO menin peaks at 15 canonical menin-MLL target loci, while EPZ5676 keeps 14, compound 10 keeps 14 and compound 11 keeps 8 - strength-matched permutation p = 0.0003, and the zero is unchanged at every flank from 0 to 100 kb, In the same deposit, the MENIN inhibitor removes DOT1L from all five DMSO DOT1L peaks at canonical target loci (p = 0.0432) while all three DOT1L inhibitors keep 5/5 - and the VTP DOT1L library still calls ACTB, GAPDH and EEF1A1, which is the internal positive control the MEN1 arm lacks, Novartis 'compound 10' - a DOT1L methyltransferase inhibitor from Perner et al. (Blood 2020), active at 100 nM, whose chemical structure is not given in the GEO deposit and whose target engagement at that dose IS demonstrated in the same series by a 5.6-fold spike-in-normalised H3K79me2 collapse by day 3, Novartis 'compound 11' - a DOT1L methyltransferase inhibitor from Perner et al. (Blood 2020), active at 100 nM, whose chemical structure is not given in the GEO deposit and whose target engagement is not measured anywhere in the series, Resolving cycle 83's threat: the MLL1 channel moves under both drug classes and is therefore not diagnostic, but the MEN1 channel separates them completely - so 'menin displacement' remains a specific readout of menin-interface engagement and the corpus must stop treating the two channels as one, The deposited dm6 spike-in table covers the eighteen H3K79me2 samples and nothing else - the ten MEN1 and DOT1L ChIPs in the same series have no fly chromatin, so every occupancy number from them is a within-sample contrast and no gain-or-loss statement on an absolute scale is available, The menin inhibitor in this head-to-head is at 50 nM and all three DOT1L inhibitors are at 100 nM - the comparison is dose-biased AGAINST the menin inhibitor - and compound 10's target engagement at exactly 100 nM is demonstrated inside the same series by a spike-in-normalised H3K79me2 collapse of 5.6-fold by day 3, The one experiment that would close this beat: MEN1 and DOT1L ChIP in RS4;11 (KMT2A::AFF1) under DMSO, a menin inhibitor and a target-engaging DOT1L inhibitor, with Drosophila spike-in and two replicates per arm - about 24 libraries, and the depositing lab already runs every component of it, The spike-in ruler this thread has been blocked on for eight cycles is deposited as a 2 MB text file, and it includes RS4;11 - the index fusion: under DOT1L inhibition the fly fraction rises 4.9x to 23.4x against matched DMSO, so H3K79me2 collapses GLOBALLY by roughly that factor, and a within-sample share would have shown almost none of it, The whole-peak-set overlap metric cannot tell a MEN1 ChIP from a DOT1L ChIP in the same untreated cells (Jaccard 0.357, inside the 0.310-0.487 range of all 45 sample pairs) - so on the same data it scores cycle 83's threat as SURVIVING while the pre-specified locus set scores it as FALLING, Under all three DOT1L inhibitors, DOT1L's fold-enrichment at canonical target loci is HIGHER than DMSO (PBX3 4.85 to 13.92, SENP6 3.37 to 9.68) - either catalytically dead DOT1L is retained or trapped on chromatin, or the reading is an artefact of a 1.35-fold spread in MACS fragment-size estimates, and this deposit cannot tell which, VTP-50469 - the preclinical menin-KMT2A interface inhibitor that almost every chromatin experiment in this field actually used, and which had no entry of its own in this graph
78. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 41778833
    Targeting Menin in T-lineage Acute Lymphoblastic Leukemia.

    Shimamoto K, et al. · Mol Cancer Ther 2026
    used by A phosphosite on a menin target predicts who responds to a menin inhibitor
79. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    Q114 pre-registration §5 (control C1) and §7 (Question A), written 2026-09-22 before any locus read

    Q114-suv39h1-at-the-locus · oncofusion-knowledge-graph 2026
    used by A pre-registered gate of the form 'signal at the top sites / median of random tiles >= 5' returns inf and PASSES on any sparse track, because the median of random 147 bp tiles in a CUT&RUN file is exactly 0.0 - the control was degenerate in three of four tracks and nothing in the output said so, In patient-derived H3K27M glioma, baseline H3K9me3 over the 7,466 bp KMT2A breakpoint cluster core sits at the 7.8th and 28.5th percentile of 59,821 width-matched CTCF-anchor windows (0.56x and 0.90x the random-tile median) - the methyl-poor-BCR pattern this graph found in epithelium and in a blood line, now in a third lineage and for the trimethyl state, Read at last, against a pre-registered class-matched null: in SU-DIPG13 and SU-DIPG36 the SUV39H1 CUT&RUN mean at the KMT2A exon-11 CTCF anchor is 0.0000 and 1.6660 - the first an EXPLICIT zero record, not missing data - and no one of the seven KMT2A intervals in either line exceeds 1.21x the genome-wide random-tile mean. The pre-registered NEGATIVE prior held., The H3K9me3 this graph argues about at the KMT2A exon-11 anchor is not deposited by SUV39H1: in SU-DIPG13 the mark sits at the 57.8th percentile of class-matched CTCF anchors while the writer reads an EXPLICIT ZERO in the same nuclei, and genome-wide the two barely track each other (Spearman 0.148, 0.078), The one experiment that would close the SUV39H1 arm of this graph's H3K9 argument: CST #8729S CUT&RUN in a CD34+ progenitor and in SEM or RS4;11, with an IgG arm and a SUV39H1-degron arm that neither existing deposit has - because after this cycle the enzyme has been measured in exactly two human cell types and neither is a blood cell, The rarest thing in the deposit, measured: across 59,821 CTCF anchors in the SAME nuclei, SUV39H1 occupancy and H3K9me3 correlate at Spearman 0.148 and 0.078 - and a quarter to two-fifths of the enzyme's own strongest sites carry no more of its mark than the median anchor, Two coupled traps in one conversion: pybigtools writes only the chromosomes the input actually touched, so a scaffold in your query set raises KeyError on one track and not another - and when the scaffolds were dropped to fix it, 71 of 59,892 anchors (0.12%) turned out to carry enough of the top signal to HALVE a control's numerator, Two more independent cell types put the KMT2A exon-11 CTCF summit within 12 bp of the coordinate this graph carries: MCF-7 at chr11:118,488,689 (5 bp away) and SK-N-SH at chr11:118,488,682 (12 bp), both score-1000 optimal-IDR peaks - and in the same tracks the anchor ranks only 72.95th of 56,171 MCF-7 anchors, so it is an unambiguous anchor and an ordinary one
80. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
    Q155 geometry re-derivation: the project's promoters.tsv against UCSC hg38 refGene, ncbiRefSeq, knownGene and Ensembl REST

    Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
    used by A promoter neighbour goes missing two ways - absent from the table, or present under a different one of its own TSSs - and both return the same clean negative, REFUTED THE SAME CYCLE: window displacement against the current annotation is 1.83%, not systematic - and the real blind spot it hid is one row per symbol at 14.69%, The AFF1 promoter window holds three genes' TSSs, not two, and the third is protein-coding - Q153's 92.5% is confirmed and its two-gene framing is not
81. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Ensembl REST lookup/id for ENSG00000163633 and ENSG00000285458, GRCh38

    Ensembl REST API, rest.ensembl.org · Ensembl 2026
    used by A promoter neighbour goes missing two ways - absent from the table, or present under a different one of its own TSSs - and both return the same clean negative
82. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
    Q155 per-sample peak scoring and promoter-matrix occupancy at the AFF1/C4orf36 locus, GSE252378

    Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
    used by A promoter neighbour goes missing two ways - absent from the table, or present under a different one of its own TSSs - and both return the same clean negative, C4orf36 - a protein-coding gene with three current TSSs spanning 43,782 bp, two of them inside or beside AFF1's promoter window, and the project's table has room for one, The fusion-biased peak at the AFF1 locus fires in 23 of 27 samples across all six partners, so it is not the fusion visiting its own partner
83. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
    The promoter table's second blind spot is not annotation drift at 1.8% - it is one row per symbol at 14.7%, and the case that motivated the question is not displaced at all

    Iqbal Lab oncofusion knowledge graph, beat Q156 (coordinator) · this project 2026
    used by A promoter neighbour goes missing two ways - absent from the table, or present under a different one of its own TSSs - and both return the same clean negative, C4orf36 - a protein-coding gene with three current TSSs spanning 43,782 bp, two of them inside or beside AFF1's promoter window, and the project's table has room for one, REFUTED THE SAME CYCLE: window displacement against the current annotation is 1.83%, not systematic - and the real blind spot it hid is one row per symbol at 14.69%, RUN, by Q156, within the hour: the whole-table displacement audit - drift is 1.83%, the real blind spot is one row per symbol at 14.69%, and the case that motivated it is not displaced, The AFF1 promoter window holds three genes' TSSs, not two, and the third is protein-coding - Q153's 92.5% is confirmed and its two-gene framing is not
84. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    UniProtKB REST API, cc_subcellular_location in JSON with evidence arrays, queried for 299 accessions on 2026-09-20

    UniProt Consortium · UniProtKB 2026
    used by A quarter of this graph's gene entries have a UniProt localization that nobody measured - and the subnuclear vocabulary UniProt does carry avoids the genes we care about most
85. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    UniProtKB REST API location values for 299 graph accessions, matched against a subnuclear/mitotic term list, 2026-09-20

    UniProt Consortium · UniProtKB 2026
    used by A quarter of this graph's gene entries have a UniProt localization that nobody measured - and the subnuclear vocabulary UniProt does carry avoids the genes we care about most
86. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
    citation-recency-fragility: publication-year distribution of cited work, split by confidence tier and by internal versus external source

    · Oncofusion knowledge graph (internal analysis) 2026
    used by A quarter of this graph's load-bearing evidence is under twelve months old, and the year field is right in a way a checker would call wrong
87. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
    analyses/q36-menin-residue-axis/run.py - contact-graph domain parse, solvent accessibility and ligand footprints over all 69 RCSB deposits referencing UniProt O00255, run 2026-09-22

    · RCSB PDB and UniProt, computed 2026
    used by A residue axis for menin, derived from coordinates: three spatial lobes, not four named ones, and the sharpest seam in the protein runs through the drug pocket
88. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    DepMap portal - bot verification interstitial served at HTTP 200 on /portal/api/download/files and /portal/api/gene/IRX1

    Broad Institute DepMap portal · depmap.org 2026
    used by A resource status has three values - verified-present, verified-absent, unreachable - and an HTTP 200 from depmap.org is a 5,175-byte bot wall titled 'DepMap - Verification'
89. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
    NCBI E-utilities esearch against db=geoprofiles - ErrorList/PhraseNotFound accompanying a Count of 0

    NCBI Entrez Programming Utilities · eutils.ncbi.nlm.nih.gov 2026
    used by A resource status has three values - verified-present, verified-absent, unreachable - and an HTTP 200 from depmap.org is a 5,175-byte bot wall titled 'DepMap - Verification'
90. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UCSC hg38 refGene table (coordinates for every interval used in q145)

    UCSC Genome Browser · UCSC 2026
    used by A reversed genomic interval returns a plausible low number and no error — the coordinate gate caught mine, HOXA is occupied as a 107 kb cluster; IRX1 is occupied as one gene, with both IRX clusters cold, IRX1 — Iroquois homeobox 1, the gene that names the HOXA-negative half of infant t(4;11) ALL
91. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42213647
    DOT1L Shapes ncPRC1-Target Gene Repression to Maintain Germinal Center B Cell Identity of Diffuse Large B cell Lymphoma.

    Gobel C, Niccolai R, Gregoricchio S, de Groot MHP, Kneefel SFK, Kreft M, et al. · Blood 2026
    used by A second disease finds the same DOT1L-to-non-canonical-PRC1 axis in the same year - and the mark moves the opposite way, H2AK119ub - the mark every PRC1.1 claim in this graph rests on, named in fifteen entries and written in none until now, H3K79 methylation is what keeps Polycomb off - and that, not gene silencing, is what a DOT1L inhibitor actually undoes, RING1 (RING1A) - the paralogue that is not RING1B, and the reason a substring search for 'RING1' in this graph returns the wrong gene, TRIM27 - the E3 ligase inside PRC1.1 that USP7 spends its time undoing, and the worst UniProt localisation record this graph has audited, USP7 - the enzyme that holds PRC1.1 together, and the only pharmacological handle on this complex that exists today
92. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Cellosaurus cell-line records: CVCL_0095 SEM, CVCL_0093 RS4;11, CVCL_0064 MV4-11 (all KMT2A-AFF1); CVCL_2119 MOLM-13, CVCL_0006 THP-1, CVCL_1609 NOMO-1 (KMT2A-MLLT3); CVCL_1866 KOPN-8 (KMT2A-MLLT1); CVCL_1418 ML-2 (KMT2A-AFDN). Gene-fusion comments fetched 2026-09-23

    · Cellosaurus (SIB) 2026
    used by A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, Of every antigen this project names, only NG2, LAMP5 and CD123 are present on all three KMT2A::AFF1 cell lines - CD19 is at 0.2 nTPM in MV4-11 and CD33 at 2.3 in RS4;11
93. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB REST, reviewed human proteins with subcellular location Cell membrane (SL-0039) plus topological-domain and lipidation features, release 2026_03, fetched 2026-09-23

    · UniProt 2026
    used by A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, LAMP5-directed therapy for KMT2A-rearranged leukaemia would meet the basal ganglia, in a disease treated in infants and routinely treated into the CNS, S1PR4 - the third parenchymally clean surface protein in the screen, a leukocyte-restricted GPCR, and not fusion-specific either, SIT1 - one of only three surface proteins in the screen that is both expressed in KMT2A::AFF1 lines and absent from normal parenchyma, and it is clean because it is leukocyte-restricted, not because it is leukaemic
94. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    Human Protein Atlas, RNA cell line nTPM (rna_celline.tsv), 1,206 human cell lines; HPA release v25.1, file Last-Modified 2025-11-05, fetched 2026-09-23

    · Human Protein Atlas 2026
    used by A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, CCNT2 / cyclin T2 - expressed in every KMT2A-rearranged line and needed by none of them, so the cyclin T split is real rather than an expression artefact, CD19 looks lower in KMT2A-rearranged patients in the RNA-seq arm of TARGET and identical in the microarray arm of the same study - so this graph should not record a CD19 deficit, CDK12 - a different kinase with a different cyclin, and in KMT2A-rearranged lines it is an anti-dependency, CDK13 - rank 171 of 17,728, equally deep in both lineage halves, and essentially unstudied in leukaemia, CDK9 - the kinase the whole SEC story runs through, which this graph held only as half of a two-protein summary entry, CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose, EZH1 - the paralogue that makes the approved EZH2 drug the wrong drug for this disease, EZH2 - the Polycomb enzyme that collaborates with the fusion instead of opposing it, and that no single-agent inhibitor can switch off, HEXIM1 - the brake pad, the only part of the reservoir anyone has deliberately raised in a KMT2A-rearranged cell, and the result was that the leukaemia grew less, HEXIM2 - the paralogue that holds the bound P-TEFb pool constant when HEXIM1 is removed, which is why single-gene experiments on this axis read flat, HPA's per-cell-line RNA file is not dead - it lives under /download/tsv/ while proteinatlas.tsv lives under /download/, and the wrong prefix returns HTTP 200 with a 117,090-byte HTML page, LAMP5 (BAD-LAMP) - a direct MLL-fusion target gene, a proposed immunotherapy antigen, and a brain protein, LARP7 - the clamp on the 3' end of 7SK, whose human loss-of-function phenotype is dwarfism and intellectual disability rather than cancer, Mediator's KMT2A signal is in its pan-essential core, CDK8 is expressed and not required at all, and MED13L is rank 9 while its paralogue MED13 is rank 13,569, MEPCE - absent from this graph until now, and a deeper pan-dependency than BRD4, Of every antigen this project names, only NG2, LAMP5 and CD123 are present on all three KMT2A::AFF1 cell lines - CD19 is at 0.2 nTPM in MV4-11 and CD33 at 2.3 in RS4;11, RBBP4 - the shared histone-binding subunit that every leukaemia line needs, and whose identical-length paralogue cannot replace it, RBBP7 - the silent paralogue that is not silent, and the reason this project's paralogue rule finally produced a positive answer, RN7SK / 7SK snRNA - the 332-nucleotide brake on elongation, and the one molecule in this beat that every dataset this project uses is blind to, Tazemetostat - an approved EZH2 inhibitor, and the drug this disease's own biology predicts will not work, The PRC2 dependency profile, lineage-split and expression-controlled - and the first time this project's paralogue rule came back positive, The transcriptional CDK with a real KMT2A-rearranged selectivity signal is CDK13, not CDK9 - and nobody has looked, What was searched, what came back empty, and what was read and not used - the P-TEFb reservoir beat's audit trail
95. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    cBioPortal, Pediatric Acute Lymphoid Leukemia - Phase II (TARGET, 2018), study all_phase2_target_2018_pub; RNA-seq RPKM and microarray mRNA profiles split on the study's own MLL_STATUS sample attribute; fetched 2026-09-23

    · cBioPortal for Cancer Genomics 2026
    used by A surfaceome-first screen for KMT2A::AFF1: 2,404 surface proteins, 169 expressed in the fusion's own cell lines, 15 fusion-enriched, 3 clean in normal tissue - and the two sets do not intersect, CD19 looks lower in KMT2A-rearranged patients in the RNA-seq arm of TARGET and identical in the microarray arm of the same study - so this graph should not record a CD19 deficit, NG2 in KMT2A-rearranged patients replicates in two independent TARGET profiles - 3 of 4 and 9 of 13 above the 90th percentile of everyone else - but it is bimodal, and the 552-fold median ratio I first computed is an artefact of a near-zero denominator, NG2 is absent from the candidate leukaemia stem-cell populations in t(4;11) ALL and 11q23 AML, so an NG2-directed therapy would debulk the disease and leave the fraction that initiates relapse, Treat a KMT2A::AFF1 patient-derived xenograft with an NG2-directed agent and read out the CD34+CD19-CD117+ fraction at nadir - does the reservoir the 2006 flow data predicts actually survive?
96. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    Abstract-only recoverability audit: how many papers cited as 'full text not read' were open-access full text

    · Oncofusion knowledge graph (internal analysis) 2026
    used by A third of the papers we said we could not read were free to read in full
97. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
    reading-queue.py - declared abstract-only, open-access full text, and low claim-to-abstract support, intersected

    · Oncofusion knowledge graph (internal tooling) 2026
    used by A third of the papers we said we could not read were free to read in full
98. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB Q8IZP0 (ABI1_HUMAN) Abl interactor 1 / e3B1, Homo sapiens, 508 aa

    UniProt Consortium · UniProtKB 2026
    used by ABI1 (Abl interactor 1 / e3B1) — WAVE-complex cytoskeletal adaptor, KMT2A partner, and an ENL-binding protein, KMT2A::ABI1 (MLL-ABI1), t(10;11)(p11.2;q23) — the partner that binds an elongation partner
99. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
    UniProtKB entry P00519 (ABL1_HUMAN): reviewed sequence, length and positional feature annotations

    UniProt Consortium · UniProtKB/Swiss-Prot 2026
    used by ABL1 — domain architecture and residue coordinates (with the 1a/1b numbering trap)
100. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     AlphaFold Protein Structure Database entry for P00519 (model AF-P00519-F1)

     Jumper J, Varadi M, et al.; EMBL-EBI/DeepMind · AlphaFold DB 2026
     used by ABL1 — domain architecture and residue coordinates (with the 1a/1b numbering trap)
101. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro member-database domain matches for UniProt P00519 (Pfam, PROSITE, SMART, CDD, CATH-Gene3D, SUPERFAMILY)

     InterPro Consortium; EMBL-EBI · InterPro 2026
     used by ABL1 — domain architecture and residue coordinates (with the 1a/1b numbering trap)
102. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     SIFTS UniProt residue-level mappings for PDB entries 1ZZP, 1OPL, 2HYY, 2ABL, 1BBZ (construct ranges read back from the PDBe mappings API)

     PDBe / SIFTS; EMBL-EBI · PDBe SIFTS 2026
     used by ABL1 — domain architecture and residue coordinates (with the 1a/1b numbering trap)
103. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     absence-claims-bounded-or-not/run.py: every absence assertion in the graph, scored for whether a date, a named source or a count bounds it

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Absence claims: 93% are unbounded in prose, but the evidence claims are 71% bounded - and the query that tested them was itself broken
104. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     control.py: positive controls showing the field-scoped AND-of-OR construction returns 0 while its own operands return 535 and 7,161

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Absence claims: 93% are unbounded in prose, but the evidence claims are 71% bounded - and the query that tested them was itself broken
105. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     NCBI E-utilities esearch, db=pubmed, executed via scripts/pubmed_safe.py; full query strings, counts and returned querytranslations persisted

     Q127 beat, oncofusion knowledge graph · analyses/q127-the-pi-dimensions-at-the-centre/bounded-queries.json 2026
     used by Absence: the tissue with the clinical signal is the one tissue nobody has put under a menin inhibitor and measured - one record in the whole literature, and it is a review, ACADM is a mitochondrial matrix beta-oxidation enzyme with an experimentally coded matrix-only localisation, recovered on 8 spectra from a nuclear-extract pulldown whose specific tail is 14.5% mitochondrial — declined, AF9's nuclear residence is chaperone-dependent, the Hsp90 site is the SAME C-terminal region the fusion keeps, and UniProt records no cytoplasmic pool at all, ALKBH5 — the second m6A eraser: a cancer-stem-cell self-renewal role elsewhere, one incidental KMT2A record, and a flat dependency row, Another fusion family did the experiment this beat wants: in FUS::ERG and CBFA2T3::GLIS2 AML, MHC class I and II are absent at diagnosis, EZH2 drives it, and an EZH2 inhibitor restores presentation, ARID1A (BAF250A) - required in the MLL-AF9 screen, present at the Myc enhancers, and flat in every selectivity measure, ARID2 (BAF200) - the PBAF subunit whose loss ACCELERATES MLL-AF9 leukaemia, and which is nonetheless required to maintain it, ATXN7L3 — the shared adaptor that actually sets H2Bub levels, and a KMT2A-selectivity hit nobody in this field has looked at, Audit of the infant presenting-feature evidence: what exists, what does not, what was rejected, what could not be read, B2M (beta-2 microglobulin) - the invariant light chain every MHC class I molecule needs, the classic class I escape route, and a gene this disease's literature has never once named, BICRA (GLTSCR1) - the ncBAF scaffold, zero PubMed records in KMT2A leukaemia, and a lineage-independent dependency in KMT2A-rearranged lines, Bounding an absence the graph already asserts: the H3K79me2 decay half-life, counted, dated and controlled - and why three of the six hits are about the drug's half-life, BRD3 - named in the founding BET paper's own mechanism and in one title since, and not once separated from BRD4 by any experiment in this disease, CBX8 — the one place in this interactome where two good papers flatly disagree, and DepMap takes a side, CD15 (Lewis-x, SSEA-1, 3-fucosyl-N-acetyllactosamine) - a sugar, not a protein: the most-named antigen in this graph that no gene encodes, CD274 (PD-L1) - named in zero entries of this graph until now, a direct MLL1-H3K4me3 target gene in another tissue, and a cell-intrinsic dependency of MLL-AF9 leukaemia-initiating cells, CD65 - a neutral glycolipid epitope (V3FucnLc8Cer), no gene, and the CD number that was split in two because its antibodies did not agree, CD65s (VIM-2, sialylated ceramide dodecasaccharide) - the myeloid antigen that, with CD15, makes NG2 a perfect predictor of KMT2A rearrangement, CDC73 / parafibromin — the PAF1C subunit that was genetically tested, and a second tumour suppressor required by the fusion, CDK4 - the other half of the class name, and it is NOT the gene this disease needs, CEP112 is a 955-residue centrosomal coiled-coil protein on 7 spectra with zero leukaemia literature — its only nuclear annotation is a single 2024 citation, and a long coiled-coil is the classic non-specific AP-MS passenger — declined, CHEK2 (CHK2) - named in every combination mechanism here and measured on its own in none of them, Children with NPM1::ALK lymphoma immunise themselves against their own fusion protein, and the antibody titre predicts relapse - while no autoantibody against any KMT2A fusion has ever been reported. The difference is not the junction; it is that ALK is an ectopic self-antigen and KMT2A and AFF1 are not., CHUK (IKK-alpha) — the only kinase that sits in BOTH NF-kB arms, which makes it the one subunit whose inhibition has an ambiguous sign in this disease, Clec12a is REQUIRED for NUP98::NSD1 AML pathogenesis - a marker turning out to be a dependency, in a different fusion family, and untested in KMT2A-r, CORRECTED: two of eleven CXXC readers HAVE been measured alongside a KMT2A fusion, and the result supports the menin-interface hypothesis, CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose, CTR9 — a PAF1C subunit the field assumes is a fusion partner, with no paper that tested it, CXXC4 (IDAX) - TET2's CXXC domain, sitting on a separate gene after an evolutionary inversion, and annotated cytoplasmic only by similarity, CXXC5 (RINF) - the one CXXC reader measurably down-regulated in KMT2A-rearranged AML, which is the only direction-of-change any family member has in this disease, DDX6 - the only protein ever proposed to do for a cell what Tat does for the virus, on the strength of one 2016 paper, DNMT3A - an independent adverse factor in adult KMT2A-rearranged AML and conspicuously absent from paediatric disease, which is the whole finding, DNTT (TdT) - a real gene the graph calls by its enzyme nickname, and TdT-negativity is 100% KMT2A-rearranged in the largest paediatric series, DOT1L's localization record is a mouse kidney paper wearing a human label - and the one image of it in a leukaemia cell is a supporting figure a title search cannot see, DPF2 (BAF45D) - a cBAF-specific reader required in MLL-AF9 cells, with one PubMed record in this disease and no selectivity signal, EBNA1BP2 (EBP2) - the DOT1L spoke with the HIGHEST stoichiometry and the WEAKEST evidence, and the case that shows why abundance is the third dimension, ELP5 (DERP6) - the bridge that joins ELP3 to ELP4, the most essential Elongator subunit in DepMap, and a gene whose cytoplasmic call rests on an overexpression experiment, ELP6 (TMEM103/C3orf75) - an Elongator subunit named after a membrane protein it is not, whose entire localisation record comes from a Drosophila paper its own symbol cannot retrieve, EP400 - the ATPase that swaps H2A.Z in; its complex is required by MLL leukaemia stem cells, and in DepMap it is one of the three genes LEAST required in KMT2A-rearranged lines, Every partner this project imports from the only MLL-fusion AP-MS interactome was pulled from a Strep-HA transgene in Jurkat T-ALL cells — so the host's own lineage factors are in the prey pool, and one of them is LEF1, EWS::FLI1 builds its own enhancers at GGAA microsatellites, and a single inherited base change that lengthens one repeat raises the fusion's binding and the patient's risk. KMT2A fusions occupy pre-existing, invariant CpG islands - which is why this disease has no germline modifier of its target set, and why its mouse models work at all., Experiment — is a KMT2A-septin fusion menin-dependent? The one-plate version, Fever and fatigue: two of the five headline signs cannot be elicited from a baby, Four YEATS antagonists exist, four public transcriptome datasets of YEATS inhibition exist, and not one paper or dataset has ever asked what happens to the AFF2- or AFF3-scaffolded complexes — and the analysis needs no wet work, FTO — highly expressed in MLL-rearranged AML, and the single flattest dependency row in this beat: 0.2% of 1,208 lines, FUT4 - the myeloid alpha-1,3-fucosyltransferase, a hub gene of the KMT2A-rearranged expression module, and the gene HGNC once officially called CD15, FUT9 - the dedicated Lewis-x synthase, structurally solved, and untouched by the KMT2A literature, GATA2 - the least KMT2A-selective gene on this beat, ranking 17,591 of 17,728, and a factor whose leukaemia biology is about stemness and chemoresistance rather than about the fusion, GLUD1 is a hexameric mitochondrial matrix dehydrogenase recovered on 11 spectra with non-zero control signal from a nuclear pulldown — and this corpus has already declined one KMT2A partner whose sole source was a cryo-EM structure of GLUD1 — declined, HCFC2 — menin's other host cell factor, present in the founding paper by gel filtration and never mentioned again, HLA-A - the class I heavy chain a menin inhibitor was actually measured on, and the human orthologue of the MHC class I molecule that protects MLL-AF9 leukaemia stem cells from NK cells, HLA-B - the class I heavy chain the bivalency screen was actually read out on, and the locus that presents the one KMT2A epitope this graph has a structure for, HLA-DR - a two-chain MHC class II heterodimer, so no single gene symbol denotes the surface marker a flow report calls HLA-DR, HLA-DRB1 - the polymorphic beta chain, one of several that can make an HLA-DR dimer, and the reason a chain cannot be inferred from a flow result, HNRNPA2B1 — the panel's strongest KMT2A-selective RNA-layer gene after PTBP1, with ZERO records against MLL or KMT2A, HOXA9 has its own nuclear bodies, and in monocytic AML they move into the eIF4E body - a subnuclear switch measured in patient marrow and never genotyped, Hypothesis: a menin inhibitor must come BEFORE an antigen-directed immunotherapy, not with it or after it - and every published experiment has already assumed this without anyone testing it, Hypothesis: CLEC12A is a cell-intrinsic dependency in KMT2A-rearranged leukaemia, not only a surface marker - and if so, antigen-negative escape is self-defeating, IBTK - the only one of the six DOT1L spokes with a reciprocal test, and its own pulldown in the same study does not find DOT1L, ICOSLG on infant t(4;11) blasts converts T cells into regulatory T cells - the one immune-evasion route measured in this exact disease, in this exact age group, IFNGR1 - the receptor the whole antigen-presentation mechanism runs through, and the one where low-dose interferon-gamma makes MLL-fusion leukaemia worse, IGF2BP2 — the paralogue co-over-expressed with IGF2BP3 in MLL/AF4 B-ALL, and the one with an inhibitor in another leukaemia, IGF2BP3 works in cytoplasmic RNA granules - the only compartment in this beat where a drug is already pointed, and nobody has looked at it in a leukaemia cell, Is the CpG address a contested resource? On current evidence, no - and the beat that went looking says so, JAK1 - recruited and stabilised by AF10 fusions, and the kinase whose downstream STAT1 the class I effect of menin inhibition explicitly does not need, KAT2A / GCN5 - the SAGA catalytic subunit that reaches this disease through its complex rather than through its own symbol, and whose only chemical tool degrades its paralogue too, KDM2A - the CXXC reader that antagonises KMT2A leukaemogenesis, and displaces wild-type MLL rather than the fusion, KHDRBS1 (Sam68) — an RNA-binding protein that holds the MLL-AF4 message DOWN, a class no dropout screen can find, KHDRBS2 (SLM1) — a counted absence: zero leukaemia records over five aliases, in a family that regulates the MLL-AF4 message, KHDRBS3 (SLM2/T-STAR) — the third STAR paralogue, also zero leukaemia records, also flat in DepMap, KMT2A::MLLT11 (MLL-AF1q), t(1;11)(q21;q23) — 28 patients, and not one functional experiment, Kotecha group, Perth: the infant KMT2A-rearranged laboratory this graph had cited once, LEF1 — the Wnt/β-catenin DNA-binding factor of the T-lineage, named as an MLL-AF4-specific interactor in a T-ALL cell line and with essentially no B-lineage or KMT2A literature, MAPK7 (ERK5) - a DOT1L partner on one 1:365 spoke of one atlas pulldown, and a kinase with real leukaemia literature that has never been connected to DOT1L, Mediator's KMT2A signal is in its pan-essential core, CDK8 is expressed and not required at all, and MED13L is rank 9 while its paralogue MED13 is rank 13,569, MEIS1's 'nucleus' is a profile prediction, not a measurement - and a homeodomain-less isoform is cytoplasmic and does the opposite job, Menin binds RPA2 directly, the contact maps to RPA2 aa 43-171 across its low-affinity ssDNA site, and three MEN1 patient missense mutations abolish it - a 2003 finding that this graph had never cited despite menin being its central node, Menin dependence of a non-elongation KMT2A fusion has never been demonstrated — the counted absence, and what the pivotal trial actually enrolled, Menin has no subnuclear compartment, and nobody has looked - the whole published record is two biochemical fractionations, both DNA-damage papers, and no microscopy at all, Menin inhibition reported to induce CLEC12A and sensitise KMT2A-r and NPM1-mutant AML to CLEC12A-directed CAR cells (2026 preprint, not peer reviewed), METTL14 — the other half of the m6A writer, with a real AML mechanism and zero KMT2A-specific records, METTL3 — the m6A writer with a clinical-grade inhibitor, an MLL-rearranged AML mechanism, and a B-lymphoid signal that goes the wrong way, MHC class I is a bivalent gene set, and a menin inhibitor derepresses it - shown in K-562, SCLC and neuroblastoma, and measured on HLA-A in primary AML, but never once in a KMT2A-rearranged cell, NASP - the soluble histone reservoir, and the only gene in this beat with a real selective difference at a pan-essentiality shallow enough to interpret, NF1 - the only tumour-suppressor loss in this disease, and it is a deletion, so a mutation panel cannot see it, NFKB1 (p105/p50) — the only menin-NF-kB subunit with a co-immunoprecipitation, and the arm the MLL leukaemia data points at, NIN (ninein) - a centrosomal scaffold recorded as a DOT1L partner under a UniProt isoform that UniProt itself calls dubious, NLRC5 - the MHC class I transactivator, and the route by which interferon activates class I when KMT2A and KMT2B are both gone, No fusion oncoprotein in any family has a published replication-protein-A measurement: 23 queries across five fusion families and four spelling variants each, every querytranslation read back, zero records - so the PI's analogy dimension for RPA has no comparator at all, and the index fusion's single locus-level result is the entire field, No infant or childhood KMT2A-rearranged cohort has REPORTED IDH status — and at least one has certainly MEASURED it, No orthogonal validation exists for any of the 110 MLL-AF4-specific partners, as far as four declared bounds can see — and the one partner the paper itself followed up, SETD2, is on the opposite list, shared by all seven fusions, No published experiment combines an AKAP95 perturbation with any menin or DOT1L inhibitor - bounded by an enumerated 92-record literature, not by a keyword zero, Non-elongation KMT2A partners considered and not written: what the searches actually returned, partner by partner, PAF1 is annotated as punctate everywhere in the nucleus EXCEPT the nucleolus - and a 2024 proximity map puts it in the nucleolus, Pallor and anaemia in infants: the number this graph quotes does not exist for babies, PARP16 - graded A- in the DOT1L pulldown, the best of the six, and a tail-anchored ER membrane protein that cannot be in the nucleoplasm, PML / TRIM19 (P29590) - the scaffold of PML nuclear bodies, co-purified with all seven MLL fusions, and never once studied as a KMT2A-fusion partner, Q48 enumeration: every antigen, identifier and query this beat ran - including what was searched and rejected, and the three identifier traps found live, RBM15 — m6A adaptor, AMKL fusion partner, and a named hit in the only KMT2A-biased RNA-protein CRISPR screen, RPA-coated single-stranded DNA is the structure that licenses ATR: ATRIP reads it, ATR is brought to the lesion, and CHK1 is activated - which is how this corpus's ATR and CHEK1 entries connect to the clamp that had no entry until now, RPA1 (RPA70) - the large subunit of replication protein A, which carries most of the ssDNA-binding folds and the N-terminal fold that hands the lesion to ATR, RPA1 and RPA2 are curated HGNC aliases of POLR1A and POLR1B, and RPA34 - a UniProt synonym of RPA2 - is an alias of POLR1G, so three of the ssDNA clamp's names point at RNA polymerase I subunits; the hazard is in symbol-based DATABASE lookups, not in PubMed, and the measured contamination says which, RPA2 (RPA32) - the middle subunit of replication protein A, whose disordered N-terminus is the phospho-switch of the replication-stress response and whose 43-171 region binds menin, RPA2 is the only compartment-compatible candidate of the six and the corpus already has a KMT2A resection thread it would fit — but it is an abundant ssDNA-binding subunit on 6 spectra in a benzonase-treated lysate with no RPA1 or RPA3 co-recovered — declined, thread flagged, RPA3 (RPA14) - the 121-residue subunit of replication protein A, which has no annotated domain, no HPA localisation call and no connection to this disease at all, RYBP - the subunit that makes a variant PRC1 complex catalytically active, named by two entries in this graph and never written, SMARCA2 (BRM) - the paralogue nobody measured, and the reason a DepMap rank cannot predict what a BRG1/BRM inhibitor does, SMARCA4 (BRG1) - a tumour suppressor by mutation and a maintenance requirement by knockdown, and in KMT2A-rearranged lines the dependency data say it is not selective at all, SMARCB1 (BAF47/INI1/SNF5) - zero entries, zero mentions and, on the evidence, correctly so: it is lost in primary AML rather than required, SMARCD1 (BAF60A) - rank 59 of 17,728, survives the lineage split, and had never been named in this graph or studied in this disease, SMARCD2 (BAF60B) - named in one KMT2A paper in the whole of PubMed, and the dependency it carries is myeloid identity, not the rearrangement, SMYD3 - a lysine methyltransferase with three claimed histone marks and one robust non-histone substrate, sitting on an HOXA9 feed-forward loop in MLL-AF9 AML and a fatty-acid-oxidation axis in CML stem cells, with zero interventional trials, SREBF1 / SREBP-1 - the weakest of the three MEN1 sterol partners, because it is simultaneously the pathway's readout, SS18 - a SWI/SNF subunit that is itself a fusion partner in another cancer, and a lineage-independent dependency in KMT2A-rearranged lines, SUPT6H (SPT6) - the histone chaperone that rides with RNA polymerase II, which puts it closer to the fusion's proposed mechanism than anything else in this beat, SWI/SNF enumeration for KMT2A-rearranged leukaemia: every subunit searched, every count dated, and what was read and rejected, SYNCRIP (hnRNP Q) — required for MLL leukaemia stem cells in one paper and a tumour suppressor deleted in T-ALL in another, The deferred KMT2A partners: what the counts actually are, what was written, what was rejected, and three numbers the previous enumeration got wrong, The H3K27me3 absence this graph keeps asserting is narrower than stated: the measurement exists, in the genetic arm of the paper the graph cites for the phenocopy, The rule-64 complex-form query that the RPA2 decline flagged as unrun has now been run, and it finds a published damage-state RPA occupancy measurement AT the KMT2A breakpoint cluster region - so two of this corpus's absence entries carry ids that are broader than the claims inside them, The six unwritten DOT1L partners are not six findings - they are nine spokes of ONE GFP-DOT1L pulldown, and the source paper already graded them, This graph recorded 'nobody has looked at AKAP95 in any KMT2A fusion' one cycle before reading the 2026 paper that did - KMT2A returns 0, MLL returns 2, and every cell-line token returns 0 because cell lines live in figures, Three CXXC-pathway activities converge on rDNA - KDM2B represses Pol I, KMT2A and SETD1A bind the whole locus, MLL-AF4 is predominantly nucleolar - and no paper contains two of them, TNF — the only NF-kB-activating ligand with real KMT2A::AFF1 B-ALL evidence, and the evidence says the blasts are RESISTANT to it, TRIM27 - the E3 ligase inside PRC1.1 that USP7 spends its time undoing, and the worst UniProt localisation record this graph has audited, TRRAP — a 3,859-residue DotCom subunit that every cell needs and no KMT2A cell needs more, Two published ways a fusion oncoprotein engages BAF - be a subunit, or recruit it with a disordered domain - and KMT2A has never been tested for either, UBE2A (RAD6A) — the E2 for H2B ubiquitination in one primary account, contradicted by another, and with no KMT2A literature at all, USP8 (UBPY) - the second deubiquitinase KMT2A partner, three patients, same UCH logic, WAC — the adaptor that hands RNF20/RNF40 to RNA polymerase II, and the only member of this axis whose knockdown alone killed MLL-rearranged ALL cells, WEE1 - the one checkpoint target with a KMT2A::AFF1-specific result, and the paper that also found the escape route, What this beat looked at and did not write: rejected candidates, with counts for both halves of every search, What this beat read and did not use: bounded absences, bounded rejections and one bounded unreadable, What this beat searched, what it rejected, and the two papers it could not read - after a fifth route rescued the third, Where LEDGF/p75 actually is: bound to chromatin through mitosis, in the nuclear interior with H3K36me3 - and never once imaged in a KMT2A-rearranged cell, Where RPA actually is: nucleoplasmic and diffuse until there is single-stranded DNA, then punctate prereplication foci at the G1/S border, damage-induced foci with RAD51, and a PML-body pool - and the subunits differ in how well any of this is measured, Which immune-evasion routes are actually documented in KMT2A-rearranged disease, and which are borrowed from AML and assumed - a ranked ledger with the bounded absences that define it, Wild-type AF10 is called nuclear on a construct and found in both fractions in tissue - and that disagreement decides whether CALM-AF10's cytoplasmic life is a gain or an inheritance, WTAP — an oncogenic m6A writer subunit in AML with ZERO records against MLL or KMT2A over six aliases, YAP1 (P46937) - the Hippo pathway's transcriptional co-activator, and a KMT2A fusion partner from a completely different functional class, YTHDC1 — the nuclear m6A reader in the MLL-METTL3-CTCF axis, with an inhibitor, and no dependency signal in KMT2A-rearranged lines, YTHDF2 — the cytoplasmic m6A reader that selectively compromises AML cancer stem cells, discordant across the lineage split here, ZC3H13 — the m6A writer subunit with the most evenly KMT2A-selective dependency signal in the panel, and almost no literature
106. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     NCBI GEO DataSets esearch and esummary, db=gds, executed 2026-09-23

     Q127 beat, oncofusion knowledge graph · GEO 2026
     used by Absence: the tissue with the clinical signal is the one tissue nobody has put under a menin inhibitor and measured - one record in the whole literature, and it is a review
107. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P11310 (ACADM_HUMAN), Medium-chain specific acyl-CoA dehydrogenase — accession, primary gene name, length and subcellular-location comment retrieved 2026-09-24

     · UniProt Consortium, EMBL-EBI/SIB/PIR 2026
     used by ACADM is a mitochondrial matrix beta-oxidation enzyme with an experimentally coded matrix-only localisation, recovered on 8 spectra from a nuclear-extract pulldown whose specific tail is 14.5% mitochondrial — declined
108. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Compartment audit of the Skucha 2018 MLL-AF4 prey classes against live UniProt, run 2026-09-24 by Q180. Script analyses/q180-is-af4-specificity-a-depth-artefact/compartment_of_the_tail.py, output compartment_of_the_tail.json; 152 of 152 accessions resolved.

     · This repository (script and JSON committed) plus UniProtKB REST 2026
     used by ACADM is a mitochondrial matrix beta-oxidation enzyme with an experimentally coded matrix-only localisation, recovered on 8 spectra from a nuclear-extract pulldown whose specific tail is 14.5% mitochondrial — declined, GLUD1 is a hexameric mitochondrial matrix dehydrogenase recovered on 11 spectra with non-zero control signal from a nuclear pulldown — and this corpus has already declined one KMT2A partner whose sole source was a cryo-EM structure of GLUD1 — declined
109. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9UKV3 (ACINU_HUMAN) - Apoptotic chromatin condensation inducer in the nucleus, Homo sapiens

     · UniProtKB 2026
     used by ACIN1 (Acinus) - a splicing/apoptosis protein whose only domain is the 35-residue module the leukaemic fusion keeps, and whose entire localization record is uncoded, ACIN1::NUTM1 - the commonest NUTM1 fusion in infants, cryptic to cytogenetics, and the only one with a human cell line
110. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41460961
     Comprehensive molecular and functional analysis of NUTM1-rearranged leukemia

     Nishimura K, Isobe T, Shigehiro T, et al. · Blood 2026
     used by ACIN1::NUTM1 - the commonest NUTM1 fusion in infants, cryptic to cytogenetics, and the only one with a human cell line, BRD9::NUTM1, t(5;15)(p15.33;q14) - one fusion is enough to make infant B-ALL in a mouse, and it is the one that behaves best in a child, NUTM1 (NUT) - a testis-restricted, fully disordered protein that writes no mark and instead switches ON a writer, and the 3' partner of the KMT2A-wild-type infant B-ALL with the opposite prognosis, NUTM1-rearranged B-ALL - the KMT2A-wild-type arm of the same infant trials, with 4-year EFS of 100% against 36-37% next door, The second ring — smaller KMT2A-fusion mechanism labs, and where the literature actually points
111. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ClinicalTrials.gov API v2 (raw, so whyStopped is present), pulled 2026-10-06 by analyses/q200-how-neighbouring-epigenetic-drug-classes-stopped-in-acute-leukaemia/pull.py

     · ClinicalTrials.gov 2026
     used by Across 21 stopped trials of menin, DOT1L, BET, CDK9 and LSD1 inhibitors in acute leukaemia, 12 (57%) were stopped for sponsor or enrolment reasons, and the only two risk-benefit stops are one molecule - failure here has mostly been structural, All three withdrawn mechanism-targeted trials admit infants and none of the 52 trials that do not has been withdrawn (exact p = 0.0032, noticed after looking), but the stated reasons are business and administrative and all three are 2025-2026 registrations, Of 62 registered menin- or DOT1L-inhibitor trials, 7 have stopped and none gave a safety reason - four were company or supply decisions - yet the same drugs carry dose-limiting differentiation syndrome and QT prolongation that the registry never mentions, Only 4 of 43 open menin- or DOT1L-inhibitor trials admit infants under one year (9.3%), and only one (TINI-2) is infant-specific, though infants are about a third of the characterised KMT2A-rearranged population, The only two trials stopped on a stated risk-benefit judgement are GSK2879552 (LSD1) studies in AML, while iadademstat, another LSD1 inhibitor, ran a KMT2A-rearranged-enriched expansion with an acceptable safety profile - so the stop looks molecule-specific, and its reason is unread
112. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41712378
     ACSL4-associated lipid metabolism is a distinct therapeutic vulnerability in KMT2A-rearranged acute myeloid leukemia.

     Schafer S, Rahimian E, Schmitz-Hubsch L, ..., Wurm AA · Cell Rep 2026
     used by ACSL4 - a CRISPR-screen-derived, functionally validated KMT2A-selective vulnerability in myeloid disease, and a gene whose loss normally protects cells from ferroptosis, Cell state in KMT2A-rearranged leukaemia is organised by lineage, not by the fusion — five independent axes split the same way, and every one of them splits AML from ALL rather than partner from partner, ELOVL1 - the fatty-acid elongase that infant MLL-AF4 ALL retains from its fetal origin, and the lipid enzyme on the opposite side of the divide from ACSL4, The lipid-dependent state — ACSL4 and polyunsaturated fatty acid utilisation in KMT2A-rearranged AML, against ELOVL1 and saturated elongation in KMT2A::AFF1 infant ALL, The myeloid/lymphoid metabolic split in KMT2A-rearranged leukaemia is set by CELL OF ORIGIN (fetal versus adult), not by lineage or by the fusion partner
113. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry O60488 (ACSL4_HUMAN), Long-chain-fatty-acid--CoA ligase 4

     UniProt Consortium · UniProt 2026
     used by ACSL4 - a CRISPR-screen-derived, functionally validated KMT2A-selective vulnerability in myeloid disease, and a gene whose loss normally protects cells from ferroptosis
114. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paperPMID 41657788
     Matrine enhances RLS3-induced ferroptosis at non-toxic doses in acute myeloid leukemia cells with MLL rearrangement.

     see PubMed record; title fetched from Europe PMC 2026-09-20 · journal per PubMed record 2026
     used by ACSL4 - a CRISPR-screen-derived, functionally validated KMT2A-selective vulnerability in myeloid disease, and a gene whose loss normally protects cells from ferroptosis
115. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     Run in coordinates: the protective CTCF site and the patient break-prone window are ONE element (this graph, ana-kmt2a-exon-11-ctcf-anchor-is-one-cohesin-bound-element, 2026-09-22), built from Ensembl REST GRCh38 and GRCh37 and the UCSC ENCODE clustered TFBS tracks

     Q105 beat, from Ensembl and UCSC/ENCODE · Oncofusion knowledge graph 2026
     used by Adjudicated: CTCF does not do opposite things at KMT2A - it does one thing, and this graph's contradiction is an untested premise about whether a break becomes a rearrangement, CTCF/cohesin occupancy at the KMT2A exon-11 anchor is both the cause of the break and the chaperone of its correct repair - so losing it gives FEWER breaks and MORE rearrangements, Run in coordinates: the protective CTCF site and the patient break-prone window are ONE element, it is cohesin-bound, and the 'functional cut point' is 3.75 kb away - two of the graph's three exon-11s were the same exon and the third was not
116. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.internal
     q196 claim-membership adjudication of the 150-symbol unwritten-gene queue

     · Oncofusion KG 2026
     used by Adjudicating the unwritten-gene queue by claim membership diverges from it at Spearman 0.36 - and then puts four acronym collisions in its own top five, so divergence was not improvement, CORRECTED: the AF4-specific interactome gap is ZERO, not one - RPA2 was declined on purpose in cycle 159 and my no-match read was the tool warning I ignored, None of the 166 leukaemia-family GWAS in the Catalog names KMT2A, MLL, 11q23 or infants, although ETV6::RUNX1, TCF3::PBX1 and hyperdiploid ALL each have their own scans - and a 987-case scan could only detect odds ratios of about 1.4-1.6 or more, The index fusion DISCARDS the SIAH degron: the junction is at AFF1 residue 347, the degron is at human 282-304, and three entries had it the other way because they measured a murine motif against an exon-numbering junction, The murine-coordinate sweep is a NEGATIVE result - zero new defects in 3,086 entries - and its actual yield was a fourth entry that had the degron right before Q191 did
117. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI E-utilities efetch, db=pubmed, id=42578372, retmode=xml

     National Center for Biotechnology Information · NCBI Entrez 2026
     used by ADJUDICATION: the SvID and the gamma-H2A.X motif are OVERLAPPING BUT NOT THE SAME SURFACE - and the premise that made the question urgent, that the SvID is distinct from the MLL cavity, is FALSE at the SvID's own boundary
118. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed esearch 2026-09-19 and Proteins API proteomics-ptm P55197; motif scan of the P55197 canonical FASTA

     · EMBL-EBI 2026
     used by AF10 fusions recruit and stabilise JAK1 - a kinase bound to the fusion with no phosphosite on the fusion, MLLT10/AF10: twenty observed phosphosites, twelve PIKK motifs, and not one kinase-site relation in the literature - a bounded absence, The KMT2A-r AML patient table numbers MLLT10 on isoform P55197-1, not the canonical sequence: S674 is S658 and S705 is S689
119. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     Localization: a partner's baseline compartment is often miscalled, and the database is the thing that is wrong

     · Oncofusion knowledge graph (internal entry) 2026
     used by AF9's nuclear residence is chaperone-dependent, the Hsp90 site is the SAME C-terminal region the fusion keeps, and UniProt records no cytoplasmic pool at all
120. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P55196 (AFAD_HUMAN), afadin / AFDN

     UniProt Consortium · UniProtKB 2026
     used by AFDN / AF6 — domain architecture and residue coordinates
121. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro entry-to-protein matches for UniProt P55196

     InterPro / EMBL-EBI · InterPro 2026
     used by AFDN / AF6 — domain architecture and residue coordinates
122. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P51825 (AFF1_HUMAN), AF4/FMR2 family member 1

     UniProt Consortium · UniProtKB 2026
     used by AFF1 / AF4 — domain architecture and residue coordinates, AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing, AFF1 as the subtraction control: the refutation is WITHDRAWN - the degron is lost, not retained, AFF1::KMT2A (AF4-MLL), the der(4) product: what it is made of, residue by residue, Bite-back: a target nominated from an imported mechanism inherits a fusion the index disease does not have, Does KMT2A::AFF1 deliver DOT1L to the AHD or displace it? A competition experiment in t(4;11) cells, Import class 2: AHD-dependent mechanism, which cannot transfer because KMT2A::AFF1 has no AHD, In KMT2A::AFF1 there is no oncogenic DOT1L unit: all H3K79 methylation is the endogenous MLLT10-directed activity, working on genes the fusion has opened, KMT2A::AFF1 has no mapped interface to DOT1L: it lacks both structurally defined recruitment modules, and the one complex-level route claimed for it is contested by reconstitution, Partner import: when a mechanism measured in one fusion is asserted about another, Repeat the H3K79 partition in a human KMT2A::AFF1 line: menin inhibition and MLLT10 knockout, side by side with a KMT2A::MLLT3 line, Separation-of-function test: does KMT2A::AFF1 need its C-terminal homology domain to dimerise?, The CRISPR-ChIP H3K79 partition is a mouse MLL-AF9 result: resolved from GEO after the full text proved unreachable
123. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro entry-to-protein matches for UniProt P51825

     InterPro / EMBL-EBI · InterPro 2026
     used by AFF1 / AF4 — domain architecture and residue coordinates, KMT2A::AFF1 oligomerises through the retained AFF1 CHD, putting the commonest fusion in the oligomerisation class
124. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Ensembl REST lookup for ENSG00000172493 (AFF1) with transcript expansion, and the UniProt cross-reference set for P51825

     Ensembl; UniProt Consortium · Ensembl / UniProtKB 2026
     used by AFF1 / AF4 — domain architecture and residue coordinates
125. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Ensembl REST transcript lookup for ENST00000395146 (AFF1-202) with exon expansion; codon ranges computed for this graph from exon coordinates and CDS boundaries

     Ensembl; analysis by Q2b-atlas2 · Ensembl REST API 2026
     used by AFF1 / AF4 — domain architecture and residue coordinates
126. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB search for UniProt P51825 returning a single entry (2LM0), plus entries 6K7P, 6KN5 and 6R80 for AFF4, and AlphaFold Protein Structure Database entries AF-P51825-F1 and AF-Q03164-F1, all resolved directly by API

     RCSB Protein Data Bank; AlphaFold Protein Structure Database · Databases 2026
     used by AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing
127. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     PubMed literature search for primary studies of AFF1/AF4 dependency in MLL-AF4 / KMT2A::AFF1 leukemia, returning no paper with that central claim

     PubMed (US National Library of Medicine), searched via the PubMed E-utilities API · PubMed 2026
     used by AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing, ARID1B (BAF250b) - the cBAF paralogue whose deletion PROMOTES MLL-AF9 leukaemogenesis, which makes it the one gene in this beat whose published KMT2A-fusion result points the wrong way for a drug target, Auditing the seven strikes before applying them: nine of ten decisions stand, four for different reasons than were given, and one clinical argument fails, BAHCC1 - a BAH-domain H3K27me3 reader that an MLL fusion binds and upregulates in mouse, and the locus carrying the two largest MLL1-N-reactive rises under a menin inhibitor in human RS4;11, FOXP1 - the forkhead repressor that licenses the pro-B to pre-B transition and blocks apoptosis with NF-kB, sitting at a 105x MLL1-N bin in a leukaemia whose defining lesion is a pro-B arrest, KDM4B (JMJD2B) - the weakest of the four H3K9me3 demethylases, one of the two erasers knocked down in the only KDM4-perturbation arm in any archive that touches this mark, RUNX2 - a direct KMT2A-chimera target and dependency in T-ALL, a patient-specific enhancer target in KMT2A::AFF1 B-ALL, and the gene a menin inhibitor represses to restore steroid sensitivity in t(4;11) infants, SATB1 - a nuclear-matrix and PML-body genome organiser with its own matrix-targeting sequence, the clearest subnuclear-localization entry in this beat, and a T-lineage factor that B cells do not need, The atlas heuristic is scoped to six baits instead of to publication scale, so it misses HuRI, OpenCell and five other proteome-wide screens - 49.5% of rows credited with a 'focused' publication have only atlas support, and all five of q178's credible interactors fail, The corpus citation index reads `PMID nnnnnnnn` in prose and never the `pmid` citation field — it sees 1,908 of 3,991 papers, undercounts 872 of the rest, and its headline single-cell absence was an artefact, ZMYND11 (BS69) - an H3.3K36me3 reader that restrains elongation, is itself the 5' partner of a chromatin-hijacking AML fusion driving HOXA/MEIS1/MYB, and whose promoter is the third-strongest MLL1-N bin in RS4;11
128. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, targets CHEMBL5724736 (AFF1), CHEMBL5724734 (AFF4), CHEMBL5725004 (ELL), CHEMBL5725179 (PAF1), CHEMBL4105954 (SETD1A) and CHEMBL3108650 (ASH2L): target_search and get_bioactivity queries

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing
129. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 41062835
     Stabilization of AFF1 by PARylation ensures transcriptional restart after DNA damage.

     Zhu F, Fu H, Zhu W, Li C, Yang M, Feng Z, Shao Y, Li Z, Guo Y, Huang M, Zhang Y, He Y, Lu Y, Wu Z, Zhou Q, Jiao H, Huang J, Zhang L, Xue Y, Lu H · Nat Chem Biol 2026
     used by AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing
130. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entry 2LM0 polymer-entity record: a single 125-residue chain annotated to both P42568 and P51825

     · RCSB Protein Data Bank 2026
     used by AFF1 AF4_int motif to AF9/ENL AHD: the index fusion's only documented elongation interface, and it competes with DOT1L, The index fusion's 3' partner has ONE observed binary interaction in IntAct and forty-two residues of solved structure - every SEC contact attributed to AFF1 is an AFF4 result wearing AFF1's name
131. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct REST ws/interaction/findInteractions/P51825 with the P-TEFb and AFF4 accessions, and UniProt PDB cross-reference intersections, fetched 2026-09-21; analyses/spoke-expansion-provenance-census/run.py

     · EMBL-EBI IntAct 2026
     used by AFF1 reaches P-TEFb only through AFF4, and the index fusion's elongation phenotype is therefore a claim about AFF4 availability, AFF1/AFF4-CAMKV - DECLINED, and it is a LABELLED REDISCOVERY: data/symbol-collisions.json already holds a CAMKV row saying BioPlex 2017 and 2021 are one programme and that CAMKV is deposited under split isoform accessions, The binary ladder for AFF1: does the index fusion's partner touch P-TEFb without AFF4, and does it touch AFF4 at all?, The index fusion's 3' partner has ONE observed binary interaction in IntAct and forty-two residues of solved structure - every SEC contact attributed to AFF1 is an AFF4 result wearing AFF1's name
132. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entry and polymer-entity records for 4IMY and 6CYT with each entity's UniProt mapping read back individually, plus UniProt PDB cross-reference intersection for Q9UHB7 against P50750 and O60563, fetched 2026-09-21

     · RCSB Protein Data Bank / UniProtKB 2026
     used by AFF1 reaches P-TEFb only through AFF4, and the index fusion's elongation phenotype is therefore a claim about AFF4 availability, The binary ladder for AFF1: does the index fusion's partner touch P-TEFb without AFF4, and does it touch AFF4 at all?
133. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct PSICQUIC query: pubid:39251607, full 19,403-row deposition (IMEx IM-30059), retrieved 2026-09-19

     · IntAct molecular interaction database (EMBL-EBI) 2026
     used by AFF1- and AFF4-containing super elongation complexes occupy physically distinct neighbourhoods in living cells, and KMT2A::AFF1 inherits the AFF1 one, AFF1-NPM1 proximity: nucleophosmin is the only bait in any published proximity experiment whose neighbourhood contains AFF1, KMT2A and menin together, AFF1-RBM15 proximity: the m6A-complex adaptor, itself a paediatric leukaemia fusion partner, has AFF1 in its BioID neighbourhood, AFF1's modern proximity neighbourhood is four RNA-binding proteins - NPM1, RBM15, SBDS, ZRANB2 - and it barely overlaps AFF4's, AGGF1 - an angiogenic G-patch protein whose BioID neighbourhood contains ELL, KMT2A, LEDGF, WDR5, KAT7 and RNA polymerase II, DDX55 - a large-subunit rRNA helicase that is nucleolar by imaging and nucleoplasmic by its own founding paper, and the narrowest RBP bait to reach KMT2A after DDX52, GPKOW - a G-patch spliceosomal protein whose BioID neighbourhood contains AFF4, ELL, ENL, CDK9, KMT2A and DOT1L, HNRNPC - the nascent-RNA packaging protein whose BioID neighbourhood contains AFF4, ELL, ENL, CDK9, cyclin T1 and KMT2A, KMT2A sits in the 14-3-3 proximity neighbourhood, and the YWHAG link reproduces across two independent studies, KMT2A's proximity promiscuity is BAIT-CLASS-SPECIFIC: it is recovered at twice its size-and-abundance prediction by RNA-binding-protein baits in a leukaemia cell and at exactly its prediction by transcription-factor baits in a kidney cell, NOT YET REPLICATED (corrected cycle 97). Within the K562 map KMT2A's 25-of-50 promiscuity is explained by size (97.1st percentile falling to 90.7th among proteins its own length) and menin is under-reached at the 23.5th, surviving an abundance control. The replication attempt was UNINFORMATIVE, not negative: eleven of twelve other proximity panels do not sample the nucleus, and the one that does shows the same shape, PLANNED: refit prey recovery on accessible lysine count instead of protein length, which separates the two mechanisms that could produce the size effect, PPIL4 - a spliceosomal cyclophilin whose BioID neighbourhood is the widest overlap with this graph's subject matter of any bait, Prey recovery in the two BioID depositions tracks SIZE and not abundance - and the interactome tail was being scored on the wrong side of the pair, because every one of the 22 partners is the BAIT and every core protein is the PREY, PSIP1-FASTKD2 declined: the bait is a mitochondrial matrix and nucleoid protein with two experimental localisations, the prey is nucleoplasmic LEDGF, and the row is one spoke-expanded BioID2 proximity in K562, Re-adjudicating the standing weak spot: proximity labelling HAS sampled AFF1, KMT2A and five leukaemia fusions - just never a KMT2A fusion, SBDS - a ribosome-maturation factor whose loss causes a leukaemia-predisposition syndrome, and whose BioID neighbourhood contains AFF1, KMT2A, CDK9, PAF1 and LEDGF, SMNDC1/SPF30 - a Tudor-domain splicing factor recovering ELL, ENL, KMT2A, PAF1 and KAT7, The 50-RBP BioID2 proximity map in K562 - the first modern proximity dataset this graph can reach, and AFF1 is in it as prey, The super elongation complex, sampled from the RNA side: ELL, ENL, CDK9 and cyclin T1 are all in the proximity record - ELL2, EAF1/2, AF9 and AF10 are at zero, Wild-type KMT2A is one of the most promiscuously labelled proteins in the proximity record - 25 of 50 baits - and menin is one of the least, at 3, ZC3H8 / Fliz1 - the last little-elongation-complex subunit without an entry, discovered in fetal liver haematopoietic progenitors, and a selective BioID2 bait that reaches KMT2A in K562, ZRANB2 - the only bait in any published proximity experiment that reaches AFF1, AFF4, ELL, MLLT1, CCNT1, KMT2A and DOT1L at once
134. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record for AFF1-AS1 (HGNC:53447), fetched via the HGNC REST service

     HUGO Gene Nomenclature Committee, rest.genenames.org · HGNC 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see
135. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UCSC Genome Browser hg38 refGene and ncbiRefSeq tracks, chr4:86,800,000-87,200,000, with NCBI Gene esummary for GeneIDs 100506746 and 4299

     UCSC Genome Browser, api.genome.ucsc.edu; NCBI Gene E-utilities · UCSC Genome Browser / NCBI Gene 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see
136. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Ensembl REST: xrefs/symbol/homo_sapiens/AFF1-AS1 and overlap/region/human/4:86915000-86945000 (gene and transcript features), GRCh38

     Ensembl REST API, rest.ensembl.org · Ensembl 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see
137. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     Q155 bounded literature search: 32 PubMed queries through scripts/pubmed_safe.py with querytranslation recorded, plus Europe PMC full-text and GEO gds searches

     Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see, No published biology exists for AFF1-AS1 under any name - and the 2026 review of non-coding RNAs in KMT2A-rearranged ALL does not mention it
138. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.reviewPMID 41358901
     The hidden regulators: Non-coding RNAs in KMT2A-rearranged acute lymphoblastic leukemia.

     Poersch MA, Rodrigues AC, Stricker PEF, Azevedo ALK, Bruschi DP, de Oliveira JC · Int J Cancer 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see, No published biology exists for AFF1-AS1 under any name - and the 2026 review of non-coding RNAs in KMT2A-rearranged ALL does not mention it
139. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     Q155 interval deconvolution over all 133 GSE252378 libraries, read remotely without downloading any bigWig

     Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
     used by AFF1-AS1 - the antisense RNA 151 bp from the index fusion partner's TSS, which Ensembl does not recognise and this project's promoter table cannot see, Strand-resolved transcription across the AFF1 / AFF1-AS1 / C4orf36 promoter - the only measurement that can say which gene the signal belongs to, The KMT2A-N signal at the AFF1 locus is bimodal, neither mode is on AFF1's TSS, and the flank test I ran first said the opposite
140. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     IntAct PSICQUIC query: identifier:P51825 AND pubid:39251607 (AFF1 proximity records from the 50-RBP BioID2 deposition, IMEx IM-30059)

     · IntAct molecular interaction database (EMBL-EBI) 2026
     used by AFF1-RBM15 proximity: the m6A-complex adaptor, itself a paediatric leukaemia fusion partner, has AFF1 in its BioID neighbourhood, AFF1's modern proximity neighbourhood is four RNA-binding proteins - NPM1, RBM15, SBDS, ZRANB2 - and it barely overlaps AFF4's
141. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NHGRI-EBI GWAS Catalog, ontology-annotated associations (release of 2026-10-05)

     · GWAS Catalog 2026
     used by AFF1, the fusion's 3-prime partner, carries 126 genome-wide-significant in-gene SNPs across 505 traits in one contiguous 201 kb block (98.9th length-matched percentile), 631 of 849 rows lipid-related and 68 red-cell indices, DOT1L carries 52 in-gene genome-wide-significant SNPs at the 97.8th length-matched percentile, led by height, pulse pressure and blood-pressure traits - a speculative cardiovascular lead for DOT1L-directed drugs, MEN1 and CDK9 each carry exactly one genome-wide-significant in-gene common variant and PSIP1 two, so GWAS says almost nothing about on-target liability - yet the revumenib label lists hyperparathyroidism, the organ MEN1 loss famously affects, None of the 166 leukaemia-family GWAS in the Catalog names KMT2A, MLL, 11q23 or infants, although ETV6::RUNX1, TCF3::PBX1 and hyperdiploid ALL each have their own scans - and a 987-case scan could only detect odds ratios of about 1.4-1.6 or more
142. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProt Q03164 (KMT2A_HUMAN), feature table

     UniProt Consortium · UniProtKB 2026
     used by AFF1::KMT2A (AF4-MLL), the der(4) product: what it is made of, residue by residue, All 46 patient phosphosites sit outside folded domains, none of LEDGF's twelve is in the clamp that holds KMT2A, and six of KMT2A's eleven are not on the fusion protein, Attribution audit 2 of 5 - post-translational modifications on KMT2A::AFF1: nothing survives at peptide level, and the one strong PTM result is explicitly about the wild-type protein, Chromocentres and pericentromeric satellite heterochromatin — no KMT2A evidence, but the AT-hook argument says there should be, Competing model 3: reciprocal status is a proxy for where the break fell, and the thing that changed is the DIRECT product, Does the sarcoma fusion bind menin? One co-immunoprecipitation with the right control settles six years of the question, Experiment: a topology-controlled binding survey of the MLL AT-hook region, plus a triple AT-hook alanine mutant in an endogenously tagged fusion, FYRN/FYRC: the clip that holds the cut reciprocal protein together, and the only fusion-adjacent interface anyone has tried to break with a decoy, In leukaemia KMT2A gives the front of the protein; in sarcoma it gives the back. Every consequence follows from that., In-silico digest of 81 modelled KMT2A fusion junctions: the peptide that would prove the fusion protein exists, KMT2A / MLL1 — the gene that breaks, KMT2A*: a menin-blind, LEDGF-blind, SET-competent KMT2A protein made from out-of-frame reciprocal alleles, LDB1::KMT2A - the clearest statement in this literature that a sarcoma fusion keeps KMT2A exons 4 to 36 and discards exons 1 to 3, Measure H3R2 and H4R3 methylation states at KMT2A-fusion targets by isomer-resolving mass spectrometry, and test the gate by writer perturbation read out on the enzyme's product - the experiment that has never been done in any fusion, Menin dependency tracks the availability of an intact menin-binding motif, in cis or in trans — not the fusion family, NUP98::KMT2A, inv(11)(p15q23) — the NUP98 fusion that contains KMT2A and is menin-inhibitor RESISTANT, Prey recovery in the two BioID depositions tracks SIZE and not abundance - and the interactome tail was being scored on the wrong side of the pair, because every one of the 22 partners is the BAIT and every core protein is the PREY, t(6;11) makes four fusion genes, not two, and the PHD/bromodomain cassette is what opens chromatin in all of them, t(9;11): the reciprocal product is made, is engineered into every genome-editing model, and has never been separately tested, Taspase1 is what turns the reciprocal product into an oncoprotein - and it is a conditional oncogene because of it, The 41 residues after the CXXC fold (1204-1244) carry a substantial part of the fusion's DNA grip, and removing them is why the structural literature's affinity is so low, The CXXC domain (aa 1147-1195, PDB 2KKF) is a chromatin address, not a compartment address — CpG mutants lose transformation but keep their subnuclear position, The infant breakpoint mutilates the reciprocal protein's first PHD finger - so 'the reciprocal product' is at least two different proteins, The KMT2A::MLLT1 chimeric protein: exactly what survives the join on each side, in residue numbers, The menin interface is a requirement of 5'-KMT2A fusions in haematopoiesis, not of KMT2A rearrangement as such - and a solid tumour has been quietly proving it for six years, The MT/BCB domain, Q03164 1147-1244: the CXXC fold plus a 41-residue basic tail nobody has measured on its own, The predictable clinical error this beat exists to prevent: a menin inhibitor given for a KMT2A-rearranged sarcoma because of the gene name, The reciprocal MLLT1::KMT2A transcript was reported in the paper that discovered KMT2A, and nobody has gone back to it in thirty-four years, The three AT-hooks (aa 169-180, 217-227, 301-309) as a localisation determinant — and the correction that MLL's hooks read DNA *structure*, not AT-richness, Which KMT2A exon carries which interface: exon 1 is the menin-binding motif, exons 5-6 are the CXXC zinc finger (computed, not recalled), Which of our phospho-sites are actually seen in patients, YAP1::KMT2A::YAP1 (YKY) - an insertional KMT2A rearrangement that keeps the CXXC zinc finger and throws the menin-binding exon away
143. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     Q155 occupancy scoring of the AFF1 promoter row at three thresholds, GSE252378 promoter matrix

     Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
     used by AFF1's own promoter window is occupied in 20 of 27 samples - top 4% of the whole table - and the call cannot be attributed to AFF1
144. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     Q155 controls: arm-by-arm scoring and the table-wide distribution of per-promoter call counts

     Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
     used by AFF1's own promoter window is occupied in 20 of 27 samples - top 4% of the whole table - and the call cannot be attributed to AFF1
145. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     Q155 corpus scan for promoter-scale statements about AFF1's own promoter

     Iqbal Lab oncofusion knowledge graph, beat Q155 · this project 2026
     used by AFF1's own promoter window is occupied in 20 of 27 samples - top 4% of the whole table - and the call cannot be attributed to AFF1
146. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI Gene esummary for GeneID 100506746 (AFF1-AS1) and 4299 (AFF1), assembly NC_000004.12

     NCBI Gene / E-utilities · NCBI Gene 2026
     used by AFF1's promoter window is 92.5% AFF1-AS1's - worse than IRX2's 89.2% - and the project's promoter table has no row for the neighbour at all
147. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     q153 audit of the 19,404-row promoter table and a repository-wide grep for AFF1-AS1

     Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
     used by AFF1's promoter window is 92.5% AFF1-AS1's - worse than IRX2's 89.2% - and the project's promoter table has no row for the neighbour at all, No gene-body-scale IRX2 measurement exists in this project - the one scale that could attribute the signal is the one nobody computed
148. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q8NCB2, retrieved 2026-09-24

     · UniProt Consortium 2026
     used by AFF1/AFF4-CAMKV - DECLINED, and it is a LABELLED REDISCOVERY: data/symbol-collisions.json already holds a CAMKV row saying BioPlex 2017 and 2021 are one programme and that CAMKV is deposited under split isoform accessions, Two of seven interactome read-queue partners should not get a gene entry, and the reasons are a pseudokinase with no nuclear pool and a membrane-skeleton protein whose two leukaemia hits are a gene-name artefact
149. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     data/symbol-collisions.json, rows 'BioPlex isoform splitting is SYSTEMATIC...', 'IntAct curates enzyme-substrate assays as interactions...' and 'SUMO1 / ubiquitin-like modifiers (a CLASS, not one symbol)'; read 2026-09-24

     · this project 2026
     used by AFF1/AFF4-CAMKV - DECLINED, and it is a LABELLED REDISCOVERY: data/symbol-collisions.json already holds a CAMKV row saying BioPlex 2017 and 2021 are one programme and that CAMKV is deposited under split isoform accessions, AFF4-SDCBP (syntenin): the only prey-only row in this census typed 'direct interaction' - a proteomic phage-display hit on AFF4's C-terminal peptide, in vitro, Every IntAct number in this project is drawn through a query that misses half the record - and what it misses is the taspase1 fragments, IntAct curates SEVEN MLL fusion proteins as first-class interactors - ENL, AFF4, AF9, GAS7, ELL, AF6 and CALM - and not the index one, KMT2A-AKAP8 (AKAP95): graded 'prey-only' by this project, and it is a direct positive modulator of MLL H3K4 methyltransferase activity from a dedicated MLL paper - ADMITTED, KMT2A-CLOCK: one paper, six IntAct records, and the bait/prey grade splits it down the middle - CLOCK falls in the discard bin and SIRT1 does not, from the same experiments, KMT2A-DRC5 - DECLINED. A ciliary dynein-regulatory-complex protein, recovered by BioID in a centrosome-cilium map, and the only one of 135 uncovered partners never mentioned anywhere in this corpus, KMT2A-HMGN5: recovered by endogenous CRISPR tagging in OpenCell, not by over-expression - a chromatin-decompacting nucleosome binder with a real reason to be near KMT2A, KMT2A-KLHL22 - DECLINED. The single source is a cryo-EM structure of KLHL22-CUL3 bound to glutamate dehydrogenase, and KMT2A is a HEK293F pull-down passenger with no nuclear rationale, Re-grade the interactome on PSI-MI method class instead of bait role - a dry experiment, and an EXECUTION of a prescription data/symbol-collisions.json already wrote down in cycle 57 and nobody has run, scripts/is-symbol-covered.py returns 'no match' for BRAF, AKT1, EGFR, CD226 and IRF2 - all five have a declined interface entry. The `interface-` prefix in its PREFIXES list can never match anything., The bait/prey grade this project uses to triage the interactome systematically discards its strongest evidence, because the strongest methods have no bait, The prey-only interactome is 36.3% written, not 0% - the census, by identity, of all 212 human prey-only partners of the six core proteins, The twelve partners with no entry are three times LESS spoke-expanded than the interactome they came from - 23.1% spoke-only against a 70.6% null, Fisher p = 0.00068, Two thirds of the curated interactome of this project's six core proteins is spoke-expanded - and the pairs this graph chose to write are significantly better than the ones it did not, What a curated interaction row actually claims: spoke expansion, and the ladder from 'was in the same tube' to 'touches'
150. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P51826 (AFF3_HUMAN) SUBCELLULAR LOCATION and TISSUE SPECIFICITY comments, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by AFF3 (LAF4) — the AFF-family scaffold of SEC-L3, and a KMT2A fusion partner in its own right
151. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 4IMY, 5JW9, 6K7P, 6KN5 and 6R80, with per-entity UniProt accessions, methods, resolutions and modelled residue ranges resolved against the RCSB APIs

     RCSB Protein Data Bank · RCSB PDB 2026
     used by AFF4 — the textbook SEC scaffold that KMT2A::AFF1 cells need LESS than other leukemias, ELL2 — domain architecture and residue coordinates, KMT2A::ELL (MLL-ELL), t(11;19)(q23;p13.1), The ELLbow: AFF4 301-351 wrapped over the ELL2 occludin domain at 2.0 A, Kd 86 nM - and no structure of ELL itself, the partner in t(11;19)
152. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, target CHEMBL5724734 (AF4/FMR2 family member 4): target_search, get_bioactivity and get_mechanism queries

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by AFF4 — the textbook SEC scaffold that KMT2A::AFF1 cells need LESS than other leukemias
153. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB entry Q9UHB7 (AFF4_HUMAN), SUBCELLULAR LOCATION comment and feature table

     UniProt Consortium · UniProtKB 2026
     used by AFF4 — the textbook SEC scaffold that KMT2A::AFF1 cells need LESS than other leukemias
154. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database, REST findInteractions on Q9UHB7, retrieved 2026-09-24 by q160 (partner BTK)

     · IntAct (EMBL-EBI) 2026
     used by AFF4-BTK - DECLINED. One osteosarcoma AP-MS row, and it is the only partner in the uncovered prey-only tail with an approved drug against it, which is exactly why the decline is written down, AFF4-CDK15: AFF4 is prey in a systematic CMGC-kinase TAP interactome - a kinase adjacent to the complex CDK9 works in, and nobody has followed it up, AFF4-FMR1 - DECLINED, and the provenance is the finding: the 'publication' behind this IntAct row is HENA, an Alzheimer's data-integration dataset, not an experiment on AFF4, AFF4-MED10: AFF4 is prey in a 2004 Mediator MudPIT purification - a lead with an independent mechanistic reason to be true, and no experiment that tests it, AFF4-SDCBP (syntenin): the only prey-only row in this census typed 'direct interaction' - a proteomic phage-display hit on AFF4's C-terminal peptide, in vitro, PSIP1-SDC1 - DECLINED as an interaction, kept for the splicing thread: nuclear syndecan-1 co-IP in a mesothelioma line recovers LEDGF among an RNA-splicing-enriched network, Re-grade the interactome on PSI-MI method class instead of bait role - a dry experiment, and an EXECUTION of a prescription data/symbol-collisions.json already wrote down in cycle 57 and nobody has run, The bait/prey grade this project uses to triage the interactome systematically discards its strongest evidence, because the strongest methods have no bait
155. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q9BTT4, subcellular location, length and tissue specificity, retrieved 2026-09-24

     · UniProt Consortium 2026
     used by AFF4-MED10: AFF4 is prey in a 2004 Mediator MudPIT purification - a lead with an independent mechanistic reason to be true, and no experiment that tests it
156. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     IntAct PSICQUIC, MITAB2.7 rows for Q03164 (KMT2A) and O00255 (MEN1), filtered on Q8NB90

     · EMBL-EBI IntAct 2026
     used by AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature, YWHAG / 14-3-3 gamma (P61981) - the only cross-laboratory reproduced KMT2A proximity partner in the curated record, and UniProt places it nowhere near the nucleus
157. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     analyses/core-complex-interactome-gap/results.json (generated 2026-09-22), all_missing and union fields, checked against the full node-id listing

     · this repository / EMBL-EBI IntAct 2026
     used by AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature, H3C1 stays rejected, and now with the two numbers the registry row lacked: its accession carries TEN gene names, and removing every substrate and ribosome symbol moves interactome coverage 43.8% -> 46.1%, In all 24 read-queue rows inspected, the candidate is the BAIT and this project's protein is the prey - nobody has ever tagged KMT2A, AFF4, menin or PSIP1 and found any of them; plus seven bounded rejections
158. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41933351
     AFG2A-related encephalopathy, expanding the neurodevelopmental and epileptic spectrum.

     Nou-Fontanet L, et al. · Orphanet J Rare Dis 2026
     used by AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature
159. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI PubMed esearch, symbol-generation recall test and bounded-absence queries with querytranslation recorded

     · NCBI E-utilities 2026
     used by AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature
160. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - Q8NB90 (AFG2A_HUMAN), ATPase family gene 2 protein homolog A

     · UniProt Knowledgebase 2026
     used by AFG2A / SPATA5 (Q8NB90) - a 55LCC replisome-proteostasis ATPase that co-purifies with KMT2A and menin, and whose CURRENT symbol cannot retrieve its own source paper or any of its function literature
161. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     IntAct PSICQUIC, MITAB2.7 rows for Q03164 (KMT2A) and O75475 (PSIP1), filtered on Q9BVQ7

     · EMBL-EBI IntAct 2026
     used by AFG2B / SPATA5L1 (Q9BVQ7) - the second 55LCC ATPase, the only read-queue member with an annotated nuclear pool, pan-essential in 95.5% of lines, and the one that touches PSIP1
162. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     analyses/core-complex-interactome-gap/results.json (generated 2026-09-22), all_missing and union fields, checked against the full node-id listing

     · this repository / EMBL-EBI IntAct 2026
     used by AFG2B / SPATA5L1 (Q9BVQ7) - the second 55LCC ATPase, the only read-queue member with an annotated nuclear pool, pan-essential in 95.5% of lines, and the one that touches PSIP1
163. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - Q9BVQ7 (AFG2B_HUMAN), ATPase family gene 2 protein homolog B

     · UniProt Knowledgebase 2026
     used by AFG2B / SPATA5L1 (Q9BVQ7) - the second 55LCC ATPase, the only read-queue member with an annotated nuclear pool, pan-essential in 95.5% of lines, and the one that touches PSIP1
164. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI PubMed esearch, symbol-generation recall test and bounded-absence query with querytranslation recorded

     · NCBI E-utilities 2026
     used by AFG2B / SPATA5L1 (Q9BVQ7) - the second 55LCC ATPase, the only read-queue member with an annotated nuclear pool, pan-essential in 95.5% of lines, and the one that touches PSIP1
165. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q5TGY3 (AHDC1_HUMAN) - Transcription factor Gibbin (AT-hook DNA-binding motif-containing protein 1)

     UniProt Consortium · UniProtKB 2026
     used by AHDC1 (Gibbin) - an AT-hook chromatin-architecture factor that is already a cofactor of a DIFFERENT sarcoma fusion oncoprotein, and whose KMT2A breakpoint lands in an untranslated exon
166. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     RefSeq NM_001197104.2 / NP_001184033.1 - Homo sapiens lysine methyltransferase 2A (KMT2A), transcript variant 1

     NCBI RefSeq · NCBI Nucleotide 2026
     used by AHDC1 (Gibbin) - an AT-hook chromatin-architecture factor that is already a cofactor of a DIFFERENT sarcoma fusion oncoprotein, and whose KMT2A breakpoint lands in an untranslated exon, C2CD3 - the centriole-elongation regulator fused to KMT2A exon 8 in an MPNST, predicted OUT OF FRAME, and the partner whose RefSeq transcript is 390 residues shorter than UniProt's canonical, CPT1A - the outer-mitochondrial-membrane fatty-acid gatekeeper that is also a lysine succinyltransferase, fused to KMT2A exon 3 in the one case that is tether WITHOUT address, DOCK10 (Zizimin-3) - the Cdc42/Rac1 exchange factor fused to KMT2A exon 8 in an undifferentiated pleomorphic sarcoma, and the only one of the seven this graph already held as an H3K79me2-marked gene, ELP4 (PAXNEB) - a required subunit of Elongator, the one partner of the seven with a genuine histone-acetyltransferase connection, and the one whose fusion supplies only its last 43 residues, RALGAPA2 (AS250) - the catalytic brake on RalA that behaves as a tumour suppressor, fused to KMT2A exon 22 in an osteosarcoma and predicted out of frame, which makes allele DISRUPTION the better hypothesis than a chimera, TMPRSS13 (MSPL) - a cell-surface serine protease fused in frame to KMT2A exon 20, the clearest NOT-APPLICABLE epigenetics call of the seven, and a symbol whose approved form has a third of its alias's recall
167. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     RefSeq NM_001029882.3 - Homo sapiens AT-hook DNA binding motif containing 1 (AHDC1), mRNA

     NCBI RefSeq · NCBI Nucleotide 2026
     used by AHDC1 (Gibbin) - an AT-hook chromatin-architecture factor that is already a cofactor of a DIFFERENT sarcoma fusion oncoprotein, and whose KMT2A breakpoint lands in an untranslated exon
168. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset
     DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan)

     Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
     used by AHDC1 (Gibbin) - an AT-hook chromatin-architecture factor that is already a cofactor of a DIFFERENT sarcoma fusion oncoprotein, and whose KMT2A breakpoint lands in an untranslated exon
169. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42610417
     A novel KMT2A::AFF1-derived fusion circRNA regulates mitochondrial metabolism in t(4;11) B-cell acute lymphoblastic leukemia.

     Tolomeo D, Bardini M, Venuto S, ..., Storlazzi CT · Haematologica 2026
     used by AK_7_3, a KMT2A::AFF1-derived fusion circRNA that holds down mitochondrial respiration and ROS in t(4;11) B-ALL - the only mitochondrial mechanism in this beat measured in the index fusion, Is the metabolic phenotype a consequence of the fusion programme or an independent dependency? The answer is that BOTH directions are published, and one of them points the other way entirely
170. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry O43823 (AKAP8_HUMAN) - sequence, feature table, subcellular location, subunit and tissue specificity; retrieved live 2026-09-24 by q166

     · UniProt Consortium 2026
     used by AKAP8 (AKAP95) - the only nuclear A-kinase anchoring protein, and the one that turned out to be required for MLL-rearranged leukaemia, AKAP95 directly enhances H3K4 mono-, di- and tri-methylation - by the KMT2B complex, and ONLY by the KMT2B complex; the paper's own discussion says the MLL1/KMT2A case was never tested, AKAP95-DPY30: a direct contact whose binding determinant two groups map to OPPOSITE ends of AKAP95, and neither has been structurally resolved, AKAP95's subnuclear addresses - nuclear matrix, nucleolus, and dynamic foci - and the finding that the famous one is the one its MLL-linked function does not need, Eight claims this corpus carries from PMID 23995757, checked one at a time against the full text: five confirmed, one scope-limited, one materially different, one not in the paper at all, KMT2A-AKAP8 (AKAP95): graded 'prey-only' by this project, and it is a direct positive modulator of MLL H3K4 methyltransferase activity from a dedicated MLL paper - ADMITTED
171. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41501053
     AKAP95 condensates regulate transcription and can be targeted in MLL-fusion driven oncogenesis.

     Yan X, Luan Y, Zhu Y, Cai Y, Ye F, Yang L, Zhu K, Hu J, Luo S, Lin J, Wang Y, Huang Y, Sun P, Yang Z, Shi B, Jiang H, et al. · Nat Commun 2026
     used by AKAP8 (AKAP95) - the only nuclear A-kinase anchoring protein, and the one that turned out to be required for MLL-rearranged leukaemia, AKAP8::ZNF384 - the same protein that MLL-rearranged leukaemia depends on is itself a rare B-ALL fusion partner, on the 5' side, donating its condensate-forming IDR, AKAP95 binds the Pol II CTD directly and nascent RNA gates the condensate - a feed-forward loop where transcription recruits the factor that sustains transcription, AKAP95 IDR1 co-condenses with the KMT2A N-terminal fragment - the one AKAP95 contact that lands inside the region every KMT2A fusion retains, AKAP95 is phenotypically silent at steady state and required only when a cell is changing state - the same shape in ATRA differentiation and in MLL-AF9 transformation, thirteen years apart, AKAP95 reaches this disease by two separate routes - through WRAD/DPY30 and through the retained KMT2A fragment - and only the second one survives the fusion, AKAP95-DPY30: a direct contact whose binding determinant two groups map to OPPOSITE ends of AKAP95, and neither has been structurally resolved, AKAP95's requirement in KMT2A-rearranged leukaemia runs entirely through IDR co-condensation with the retained fragment, and not at all through DPY30 - so a DPY30-binding-dead AKAP95 will fully rescue, AKAP95's subnuclear addresses - nuclear matrix, nucleolus, and dynamic foci - and the finding that the famous one is the one its MLL-linked function does not need, JD-PI95 - a genetically encoded bridging peptide that drags AKAP95 to HSP70 and dissolves its condensates, with in vivo activity in THP-1, No published experiment combines an AKAP95 perturbation with any menin or DOT1L inhibitor - bounded by an enumerated 92-record literature, not by a keyword zero, Separate AKAP95's two routes into the KMT2A axis - starting with one transfection that re-reads a thirteen-year-old control, The strongest published case that condensates are load-bearing in KMT2A-fusion leukemia — and it is AKAP95, not AFF1, This graph recorded 'nobody has looked at AKAP95 in any KMT2A fusion' one cycle before reading the 2026 paper that did - KMT2A returns 0, MLL returns 2, and every cell-line token returns 0 because cell lines live in figures, Where KMT2A::MLLT3 sits in the nucleus: two papers, both overexpression, both from before 2005 - and the fusion carries no AF9 import signal
172. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42045218
     Dynamic partitioning of a critical elongation factor between LEC and SEC regulates cellular snRNA and proliferation-related mRNA transcription.

     Nandy A, et al. · Nat Commun 2026
     used by AKT and ATM push ELL between two complexes: K355 acetylation sends it to SEC, ATM phosphorylation sends it to LEC, Anti-ELL antibodies carry a confound the other partners do not: two paralogues and a protein whose abundance is actively regulated
173. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     analyses/q109-is-the-serine-collapse-downstream-of-atf4-or-parallel - pre-registered panel over GSE127506+GSE127508, GSE63664 (every probe, own baseline) and GSE260456

     · this repository 2026
     used by ALDH1L2 (Q3SY69) - the second-largest fall in the serine/one-carbon panel (-3.40), named nowhere in this graph, zero of 1,258 DepMap lines require it, and a tumour SUPPRESSOR where it has been studied, ATF3 (P18847) - the one cascade member the DOT1L inhibitor hits HARDER than the menin inhibitor, and a transcriptional REPRESSOR being counted as a stress output, ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor, ATF4's own transcript does NOT move under either menin inhibitor - and the whole ISR arm is reproduced by a DOT1L inhibitor, so it is not a menin-interface readout, ATF5 / ATFX (Q9Y2D1) - the most reproducible repression in the earlier dataset, and the three probe values the graph quotes are SEM, not RS4;11, CBS (P35520) - the largest fall in the whole panel (-4.07), a token that collides with Cajal bodies AND CTCF binding sites, and a germline phenotype whose hallmark is thrombosis, CHAC1 / BOTCH (Q9BUX1) - the largest single effect in the whole menin-inhibitor response (-5.11 log2) and a glutathione-degrading pro-apoptotic enzyme, named nowhere in this graph until now, CTH (P32929) - gene_exact returns TWO reviewed records and the second is VSIG2, a third instance of rule 62 on the same symbol the Ensembl resolver already got wrong, DDIT3 / CHOP (P35638) - the cascade's middle node, and the symbol whose two reviewed human records and 16,893-record alias are the worst identifier trap in this beat, EIF2AK3 / PERK (Q9NZJ5) - the kinase whose transcript is flat under both inhibitors, whose arm IS functionally required in MLL-AF9 leukaemia stem cells, and whose common name collides with phospho-ERK, EIF2AK4 / GCN2 (Q9P2K8) - the amino-acid-sensing eIF2alpha kinase: transcript flat, and NOTHING has ever been published about it in any KMT2A-rearranged system under either name, GPT2 (Q8TD30) - the EARLIEST-moving and third-most menin-selective gene in the serine arm, and the one REH cannot control because it is below the array floor there, IKZF3 / Aiolos (Q9UKT9) - the B-lymphoid transcription factor that goes UP under both inhibitors, regulates BCL2, and is a degradable drug target, Menin inhibition collapses the serine/one-carbon/ASNS programme in KMT2A::AFF1 cells - PHGDH -3.24, PSAT1 -2.40, ASNS -1.72, flat in a wild-type line - so it should re-sensitise KMT2A-rearranged ALL to asparaginase, a combination PubMed has zero records for, MI-2-2 was doing a weak version of the same thing - 5.8x enriched overlap with the potent inhibitor, 75% concordant - and the selectivity claim it left behind gets STRONGER when low-expressed probes are filtered out, 9.6x to 22.1x, MTHFD1 (P11586) - the inverse case: the deepest leukaemia dependency in the serine/one-carbon panel, and the drug does not touch it, MTHFD2 (P13995) - the one panel member with a published MLL-AF9 in-vivo survival result, and the serine gene that behaves exactly like an ISR gene, One blot: phospho-eIF2alpha and ATF4 protein in RS4;11 under VTP-50469, with a dose-response and a DOT1L comparator arm, PHGDH (O43175) - the largest menin-SELECTIVE fall in the whole serine panel, and the least essential gene in it, PPP1R15A / GADD34 (O75807) - the eIF2alpha phosphatase adaptor, and KMT2A fusion proteins were shown to BIND it and block its apoptosis in 1999, a paper this graph has never cited, PSAT1 (Q9Y617) - a DIRECT ATF4 target, and the gene that shows the brief's menin-selectivity claim was PHGDH's and not the pathway's, PSPH (P78330) - the third serine-synthesis enzyme, high in Ewing sarcoma alongside PHGDH and PSAT1, and the one that does NOT fall in leukaemia, Q109 enumeration: the ten panel members that got no entry, the measurement that declined each, the papers read and not used, and four identifier traps measured on this panel, SESN2 / sestrin-2 (P58004) - the ISR's exit into mTORC1: a leucine sensor that falls 1.4 log2 under either inhibitor, and the one panel member with a real structural record, SGK1 / SGK (O00141) - a MENIN-SELECTIVE upregulated kinase, a druggable AGC family member, and the panel's worst platform disagreement, SHMT2 (P34897) - the mitochondrial one-carbon step, and the paralogue that rises while it falls is why its loss may not matter, SOCS2 (O14508) - a MENIN-INHIBITOR-SELECTIVE member of the signature, a cullin-5 E3 substrate receptor, and its likeliest upstream regulator is MEF2C rather than ATF4, STC2 / stanniocalcin-2 (O76061) - the second largest effect in the signature (-4.40 log2), a SECRETED glycoprotein, and its localisation rests on curator inference alone, The 24 genes that respond to BOTH menin inhibitors in RS4;11 are led by the ATF4 stress programme - CHAC1 -5.11, STC2 -4.40, TRIB3 -3.10 - and CHAC1 and TRIB3 are named zero times in this graph, The ATF4/ISR collapse under menin inhibitors reports loss of the KMT2A-fusion transcriptional programme by ANY route, not engagement of the menin interface - and the loss is of ATF4 activity, not ATF4, The experiment that refutes cycle 55's headline was already public: VTP-50469 collapses MEIS1 32-fold in the same line where MI-2-2 left it untouched - and HOXA9 was never actually measured by either dataset, The one experiment that settles this beat: dose-matched menin and DOT1L arms in RS4;11 with H3K4me3 ChIP at PHGDH, plus 13C-serine flux - and it answers the cycle-58 question at the same time, The serine/one-carbon arm is NOT a pure ATF4 output: it is menin-selective where the ISR arm is compound-independent (median margin -0.55 vs -0.02, p=0.011), and it is ABSENT in KMT2A::MLLT3 AML where the ISR arm is present, The serine/one-carbon programme has a SECOND, chromatin-proximal menin input that is not ATF4 - which is why PHGDH is menin-selective where CHAC1 is not, and why the arm is absent in KMT2A::MLLT3, TRIB3 / TRB3 (Q96RU7) - a pseudokinase that is the ISR's own brake, falls 3.1 log2 under either inhibitor, and has a published physical link to the MLL complex in the wrong tissue
174. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41922744
     ALDH1L2 regulates reactive oxygen species and acinar-to-ductal metaplasia in the pancreas.

     Hennequart M, Mervant L, Stockis J, et al., Vousden KH · Nat Metab 2026
     used by ALDH1L2 (Q3SY69) - the second-largest fall in the serine/one-carbon panel (-3.40), named nowhere in this graph, zero of 1,258 DepMap lines require it, and a tumour SUPPRESSOR where it has been studied
175. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q3SY69 (AL1L2_HUMAN), Mitochondrial 10-formyltetrahydrofolate dehydrogenase

     UniProt Consortium · UniProt 2026
     used by ALDH1L2 (Q3SY69) - the second-largest fall in the serine/one-carbon panel (-3.40), named nowhere in this graph, zero of 1,258 DepMap lines require it, and a tumour SUPPRESSOR where it has been studied
176. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI E-utilities esearch against PubMed, 2026-09-22, fusion operand set (KMT2A[tiab] OR MLL[tiab] OR "MLL-AF4"[tiab] OR "KMT2A::AFF1"[tiab] OR "MLL-rearranged"[tiab] OR "t(4;11)"[tiab] OR "MLL-AF9"[tiab] OR "MLL-ENL"[tiab]), each term separately tagged and every cytogenetic term quoted

     · PubMed 2026
     used by ALDH1L2 (Q3SY69) - the second-largest fall in the serine/one-carbon panel (-3.40), named nowhere in this graph, zero of 1,258 DepMap lines require it, and a tumour SUPPRESSOR where it has been studied
177. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     UniProtKB Q9UM73 (ALK_HUMAN): sequence length, domain, zinc-finger, DNA-binding, repeat, motif and region features, and PDB cross-references

     UniProt Consortium · UniProtKB 2026
     used by ALK (anaplastic lymphoma kinase) domain architecture
178. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     The RNA layer in KMT2A-rearranged leukaemia lines: panel-wide selectivity scan and IGF2BP family co-dependency test

     · Oncofusion knowledge graph (internal analysis) 2026
     used by ALKBH5 — the second m6A eraser: a cancer-stem-cell self-renewal role elsewhere, one incidental KMT2A record, and a flat dependency row, FTO — highly expressed in MLL-rearranged AML, and the single flattest dependency row in this beat: 0.2% of 1,208 lines, HNRNPA2B1 — the panel's strongest KMT2A-selective RNA-layer gene after PTBP1, with ZERO records against MLL or KMT2A, Hypothesis: IGF2BP2 and IGF2BP3 buffer each other in KMT2A::AFF1 leukaemia, so single-gene CRISPR under-reports both, I3IN-002 — the first small-molecule IGF2BP3 inhibitor with in vivo activity in an MLL-Af4 model, IGF2BP1 — the IMP paralogue that belongs to a DIFFERENT fusion, and the cleanest control the family offers, IGF2BP2 — the paralogue co-over-expressed with IGF2BP3 in MLL/AF4 B-ALL, and the one with an inhibitor in another leukaemia, KHDRBS1 (Sam68) — an RNA-binding protein that holds the MLL-AF4 message DOWN, a class no dropout screen can find, KHDRBS2 (SLM1) — a counted absence: zero leukaemia records over five aliases, in a family that regulates the MLL-AF4 message, KHDRBS3 (SLM2/T-STAR) — the third STAR paralogue, also zero leukaemia records, also flat in DepMap, METTL14 — the other half of the m6A writer, with a real AML mechanism and zero KMT2A-specific records, METTL3 — the m6A writer with a clinical-grade inhibitor, an MLL-rearranged AML mechanism, and a B-lymphoid signal that goes the wrong way, MSI2 (Musashi-2) — sustains the MLL stem-cell programme in mouse, and the myeloid DepMap arm disagrees in sign, PTBP1 — the strongest RNA-layer dependency in KMT2A-rearranged lines, found by a domain-focused screen and missed by this beat's own gene panel, RBM15 — m6A adaptor, AMKL fusion partner, and a named hit in the only KMT2A-biased RNA-protein CRISPR screen, RBM15B — the bridge from the DOT1L mark to the RNA layer: it reads H3K79me2 and steers m6A to start codons, RNA-binding proteins set how much KMT2A::AFF1 protein a cell makes from its own fusion transcript, SYNCRIP (hnRNP Q) — required for MLL leukaemia stem cells in one paper and a tumour suppressor deleted in T-ALL in another, The decisive experiment: acute degron removal of IGF2BP3, IGF2BP2 and both, read on self-renewal and xenograft initiation rather than bulk dropout, The RNA layer scored for KMT2A-selectivity across 79 genes, lineage-split — PTBP1 is the only menin-shaped hit, and the IGF2BP family shows no co-dependency, WTAP — an oncogenic m6A writer subunit in AML with ZERO records against MLL or KMT2A over six aliases, YTHDC1 — the nuclear m6A reader in the MLL-METTL3-CTCF axis, with an inhibitor, and no dependency signal in KMT2A-rearranged lines, YTHDF2 — the cytoplasmic m6A reader that selectively compromises AML cancer stem cells, discordant across the lineage split here, ZC3H13 — the m6A writer subunit with the most evenly KMT2A-selective dependency signal in the panel, and almost no literature
179. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q6P6C2 (ALKB5_HUMAN), human ALKBH5, 394 aa

     · UniProt Consortium 2026
     used by ALKBH5 — the second m6A eraser: a cancer-stem-cell self-renewal role elsewhere, one incidental KMT2A record, and a flat dependency row
180. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     GSE210480, per-sample supplementary file inventory (NCBI GEO)

     · NCBI Gene Expression Omnibus 2026
     used by All 39 bounded absences in this graph, classified by WHAT BOUNDS THEM: 23 rest on an abstract-level literature search, 7 on a threshold or a query filter, and only 2 on reading the primary source, H3K9me1 - the only H3K9 state whose level at the KMT2A breakpoint cluster region has ever been measured, and it is LOW there until KDM3B is removed, H3K9me2 - named sixty times in this graph with no entry of its own, the mark that defines the lamina-associated domains the fusion's targets sit beside, and the only H3K9 methylation state ever measured at a KMT2A fusion's own binding sites, ONE GEO series can carry TWO different chrom tables: GSE210480's H3K9 bigWigs declare hg19 + exactly 100,000 bp but its CTCF bigWigs declare EXACTLY hg19 with no padding - so a build check run on one file does not license the rest of the series, Re-analyse the eight unreported H3K9me3 samples in GSE210480 at the KMT2A breakpoint cluster region - compute only, and it decides whether the rearrangement brake is mark-specific, The anchor-matched ranking, run: the KMT2A exon-11 CTCF anchor is rank 58 of 1,625 anchors matched on baseline strength - the 3.6th percentile - and the matched MEDIAN anchor also falls to 0.63x, so 'CTCF falls only at the KMT2A anchor' is refuted while 'the KMT2A anchor loses more than most' survives weakly, The one experiment that would separate the H3K9me1/2 rearrangement brake from H3K9me3 was run, deposited as eight public samples, and reported nowhere in the paper
181. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     ClinicalTrials.gov study record NCT05631574 - Study of Covalent Menin Inhibitor BMF-219 in Adult Patients With KRAS Driven Non-Small Cell Lung Cancer, Pancreatic Cancer and Colorectal Cancer (COVALENT-102)

     · ClinicalTrials.gov 2026
     used by All 39 bounded absences in this graph, classified by WHAT BOUNDS THEM: 23 rest on an abstract-level literature search, 7 on a threshold or a query filter, and only 2 on reading the primary source, Icovamenib (BMF-219) — the covalent menin inhibitor that left oncology for diabetes, NCT05631574 (COVALENT-102): the third solid-tumour menin-inhibitor trial, missing from this graph because its drug is called BMF-219 - terminated after 13 patients for a strategic reason, explicitly NOT for efficacy or safety
182. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ENCODE portal target record /targets/BARD1-human/ (HTTP 200) versus /targets/SUV39H1-human/ (HTTP 404)

     · ENCODE Project 2026
     used by All 39 bounded absences in this graph, classified by WHAT BOUNDS THEM: 23 rest on an abstract-level literature search, 7 on a threshold or a query filter, and only 2 on reading the primary source, The GEO sweep that q110 named as free and did not run: SUV39H1 chromatin occupancy HAS been measured in two human cell types, six samples, hg38 bigWigs already deposited - so the writer half of the H3K9me3 absence is measured-not-analysed, not never-measured
183. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     PRIDE Archive project search, keyword MLL-AF4: PXD043920, PXD041875, PXD009338

     · EMBL-EBI PRIDE Archive 2026
     used by All 39 bounded absences in this graph, classified by WHAT BOUNDS THEM: 23 rest on an abstract-level literature search, 7 on a threshold or a query filter, and only 2 on reading the primary source, The junction-peptide absence was a claim about mass-spectrometry datasets bounded by a PubMed abstract search: PRIDE holds three human MLL-AF4 proteomics projects, so measured-not-searched is now named rather than asserted
184. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     Comparator-arm screen for pathway-equivalent lesions across all DepMap leukaemia lines

     · Oncofusion knowledge graph; Cellosaurus + DepMap via Open Targets 2026
     used by All 46 comparator lines screened: the two we tripped over are the only two we can find, FLT3 / CD135 — the strongest non-tether co-dependency, and a target the field may have written off on a badly executed trial, The assumption the whole project rests on, tested at last: menin dependency is fusion-specific (run), The check I called decisive was not: the haem signal is still open (run)
185. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     Mapping 46 patient-measured phosphosites onto UniProt domain architecture across ten KMT2A-complex proteins

     · Oncofusion knowledge graph (internal analysis) 2026
     used by All 46 patient phosphosites sit outside folded domains, none of LEDGF's twelve is in the clamp that holds KMT2A, and six of KMT2A's eleven are not on the fusion protein
186. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB entry O75475 (PSIP1_HUMAN), feature table including the integrase-binding domain

     · UniProt Consortium 2026
     used by All 46 patient phosphosites sit outside folded domains, none of LEDGF's twelve is in the clamp that holds KMT2A, and six of KMT2A's eleven are not on the fusion protein, PLANNED: run the size-matched promiscuity correction for PSIP1 in the K562 BioID2 map, the way it was run for KMT2A - fully specified here so the next cycle executes rather than redesigns it
187. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     interactome-bait-versus-prey/worklist.py: bait-supported partners of the core complex cross-referenced against this graph's gene entries

     · Oncofusion knowledge graph (internal analysis) 2026
     used by All 56 core-complex partners that anyone deliberately fished for already have an entry: the unwritten interactome is 100% by-product
188. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database via PSICQUIC, MITAB 2.7 records with experimental roles for the six core-complex proteins

     · EMBL-EBI IntAct 2026
     used by All 56 core-complex partners that anyone deliberately fished for already have an entry: the unwritten interactome is 100% by-product, CORRECTED SAME CYCLE: the replication test was UNINFORMATIVE, not negative. Eleven of the twelve proximity panels do not sample the nucleus - they are centrosome, RTK, kinase, ion-channel, CFTR and mitochondrial baits - so KMT2A's 0th percentile in three of them measures panel composition, not promiscuity. In the one transcription-factor panel KMT2A is at the 71.7th and menin is absent, which is DIRECTIONALLY CONSISTENT with the K562 result at n=1, Re-measured against IntAct the interactome is 103/217 (47.5%) covered, up from 41.0%, and the read queue is now ZERO - no partner with two publications or two baits lacks an entry, so the instrument's own ranking criterion has stopped selecting anything. It was also persisting the list of partners that DO have entries while counting and discarding the 105 that do not, which is the exact defect cycle 57 fixed for the other list, The IntAct interactome gap is closed - all 135 'unwritten' interactors are mentioned and 60% carry an explicit decision - while 95 of the 110 partners unique to the INDEX fusion have never been mentioned at all, The interactome tail is orderable after all, just not by evidence: 62 of the 96 unwritten partners are already DISCUSSED somewhere in this corpus and need only promoting to entries, while 34 have never been read at all. Two cycles looked for a priority order in evidence strength and found none, because the ordering that exists is over how much work is left, not over how good the evidence is, Two-thirds of this graph's core-complex interaction evidence comes from experiments where our protein was not the bait, and LEDGF has one bait-supported partner out of thirty-six
189. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     NCBI GEO GSE260742 - per-sample SOFT records, all 158 samples (targ=gsm, form=text, view=brief)

     · NCBI Gene Expression Omnibus 2026
     used by All four input samples explicitly titled to the MLLAF9 ChIP arm of GSE260742 carry supplementary_file = NONE, so no input-referenced peak call is possible from the processed deposit - and the only deposited batch-1 inputs are two BAMs filed under a different arm's title with the same cell line and the same F3 filename prefix
190. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     NCBI GEO GSM8123413 and GSM8123414 - NTvKO Control DMSO input1/input2, files GSM8123413_ChIP_INPUT1_F3_..filt.bam and GSM8123414_ChIP_INPUT2_F3_..filt.bam

     · NCBI Gene Expression Omnibus 2026
     used by All four input samples explicitly titled to the MLLAF9 ChIP arm of GSE260742 carry supplementary_file = NONE, so no input-referenced peak call is possible from the processed deposit - and the only deposited batch-1 inputs are two BAMs filed under a different arm's title with the same cell line and the same F3 filename prefix
191. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database, REST findInteractions on O00255, retrieved 2026-09-23 by Q128

     · IntAct (EMBL-EBI) 2026
     used by All nine unwritten MEN1 partners trace to exactly two publications, neither of which mentions menin outside an antibody catalogue row, Is menin an SCF-FBXW7 substrate? A one-week immunoblot and a free desk scan that settle Arm A, MEN1-AKT1 is ONE BRET well from a mutation screen, for the E17K allele only - declined, MEN1-BRAF V600E: four methods and an endogenous menin lane - the only one of the nine that is a real observation, and it is still a property of the allele, MEN1-CD226 is two proteome-scale AP-MS maps, both spoke-expanded, both in cell lines that do not express CD226 - declined, MEN1-EGFR is ONE BRET well from the extended-driver arm of a mutation screen, for L858R only - declined, MEN1-FBXW7 is four alleles of ONE BRET assay, all typed 'mutation with no effect' - declined, and it is the most insidious class, MEN1-PTEN is ONE BRET well from the extended-driver arm, for R173C only - declined, MEN1-SMAD4 is four alleles of ONE BRET assay, all 'mutation with no effect' - declined, and the real menin-TGF-beta literature is elsewhere, MEN1-SPOP is four alleles of ONE BRET assay, all 'mutation with no effect' - declined, with a nuclear-speckle confound, MEN1-ZC3HC1 is one spoke-expanded BioPlex row in HCT116, and UniProt records ZC3HC1 weakly or not expressed in leukocytes - declined, No menin interaction outside the chromatin complex has ever been tested in a haematopoietic cell - 0 of 9, and the compartment is the wrong one in every case, Three of menin's nine declined partners are E3 ligase substrate receptors - either menin turnover is ubiquitin-controlled, or substrate receptors are this beat's artefact signature
192. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     analyses/core-complex-interactome-gap/run.py, function load_graph(), read together with a word-boundary full-text scan of data/graph/nodes/*/*.json run 2026-09-23

     · oncofusion-knowledge-graph (this repository) 2026
     used by All nine unwritten MEN1 partners trace to exactly two publications, neither of which mentions menin outside an antibody catalogue row, MEN1-PIAS4 - DECLINED. The single supporting publication is a PARP-1 heat-shock sumoylation study in which menin is never named, and the row is a HeLa co-IP passenger, Menin is the best-observed of the six core proteins and the only one whose interval separates from the rest - the interface the programme is built on is the one the record actually supports, No BRCA1, BARD1, EXO1, DNA2, HP1-gamma or KDM5A has ever been localised at the KMT2A locus, in any cell type: 21 queries, every querytranslation read back, 0 on-topic records - and the absence splits THREE ways by class, which is what decides what it costs to close, The gap instrument's 'not mentioned anywhere in the corpus' reads only ids, titles and aliases - it was wrong about 9 of 9 this cycle, The interactome instrument is migrated to IntAct REST: 277 partners, 118 written, 42.6% - the denominator grew by 60 and coverage fell 3.5 points, which is the instrument getting honest, The menin arm and the CTCF arm are not two stories: menin/SETD2/H3K36me3/LEDGF licenses CtIP short-range resection and CTCF/H3K9me/HP1-gamma licenses BRCA1/BARD1-EXO1/DNA2 long-range resection, which are the two consecutive halves of ONE reaction - so a menin inhibitor and an H3K9-axis inhibitor are acting on the same pathway at different steps
193. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41145673
     CRISPR/Cas9-mediated t(4;11) translocation in human hematopoietic stem/precursor cells demonstrates plasticity to differentiate into either the myeloid or lymphoid lineage.

     Benz T, Larghero P, Meyer C, et al. · Leukemia 2026
     used by An anti-MLL-C antibody in a t(4;11) cell is looking at two proteins at once, Competing model 1: the reciprocal product starts the leukaemia and then leaves, Do different breaks give different biology? A verdict on partner identity vs breakpoint position, Settle the WRAD question: tagged co-IP plus degron-timed proximity across three fusion classes, t(9;11): the reciprocal product is made, is engineered into every genome-editing model, and has never been separately tested, The reciprocal fusion (AFF1::KMT2A and friends): a second oncoprotein that opens chromatin for the first
194. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed esearch counts, querytranslations and a zero-returning keyword search for methylarginine antibody specificity, run 2026-09-22

     · NCBI PubMed 2026
     used by An assay that cannot separate me2a from me2s cannot test the arginine switch - the two states are isobaric and predict OPPOSITE things, so a non-discriminating readout returns no information rather than an imprecise answer
195. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     ENCODE portal type=Experiment&target.label=<SYM> versus ChIP-Atlas hg38 target tables: SUV39H1 0 / 4, SETDB1 5 / 14, CBX5 4 / 29

     · ENCODE Project and ChIP-Atlas / DBCLS 2026
     used by An ENCODE-only archive bound systematically undercounts chromatin-modifying enzymes, because ENCODE's target catalogue is transcription-factor-shaped - measured 3-for-3 on the H3K9 machinery, with ratios of infinity, 2.8x and 7.3x
196. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     concept-q146-a-jaccard-is-not-a-result-until-its-null-knows-that-active-promoters-cluster (this graph)

     Iqbal Lab oncofusion knowledge graph, beat Q152 · this project 2026
     used by An independence null OVERSTATES partner-exclusive recurrence 16-fold here - it is not a lower bound, and the brief that dispatched this beat said it was
197. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     GTEx v8 median gene expression, CSPG4 (ENSG00000173546.7), 54 tissues; Artery_Tibial 246.21 TPM, median 10.9 TPM; fetched 2026-09-23

     · GTEx Portal 2026
     used by An NG2-directed therapy meets pericytes - and unlike B-cell aplasia there is no replacement therapy for a damaged vascular wall, NG2/CSPG4 is the one antigen in the whole surfaceome that KMT2A::AFF1 itself drives - and it is a pericyte and vascular protein, at 124.5 nTPM in normal blood vessel and 246 TPM in tibial artery
198. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     Q118 re-analysis of GSE127507, instrument A: the deposited series-level ROSE TSS-associated signal table, 21,461 windows x 18 samples; analyses/q118-menin-at-the-human-index-fusion/results-instrumentA.json and results-instrumentA-addendum.json, run 2026-09-23

     · Iqbal Lab oncofusion knowledge graph 2026
     used by An occupancy-matched floor built from ~8,600 mid-occupancy promoter windows per track closes a gap Q117 named in its own memo - and it was necessary, because GAPDH and ACTB GAIN share in two of three human lines while Q117 measured them LOSING share in mouse, At the same bins in the same cells, menin's share ratio is 0.20-0.26 and the MLL1 N-terminus antibody's is 0.80-0.91 - so a shared-epitope MLL1-N ChIP under-reports the menin inhibitor's occupancy effect by 3.5 to 4.0-fold, and the number is now measured rather than asserted, Fifty-three 2 kb bins gain MLL1-N-reactive signal under VTP-50469 in RS4;11, and the same 53 coordinates gain 1.77-fold in MOLM13 against a class-matched null of 1.01 with the MOLM13 input flat at 1.01 - so there IS a small, reproducible, input-controlled redistribution, and it carries no CpG-island signature, GSE127507 ran a Drosophila spike-in, measured it, and deliberately did not use it - so the VTP-50469 founding dataset carries no recoverable scale factor either, and the depositors say so in the sample record, HOXA9 is where menin leaves most completely (share ratio 0.14-0.23 over floor) and MLL1-N-reactive material stays most completely (0.74-1.02 over floor), in all three MLL-rearranged lines - so the menin inhibitor decouples menin from the HOXA9 locus without clearing the MLL1 N-terminus from it, MEF2C loses occupancy under VTP-50469 in all three human MLL-rearranged lines (MLL1-N 0.18 / 0.25 / 0.04 over floor), which resolves Q117's one locus-level disagreement with the published account in the paper's favour and localises the disagreement to the murine system, ML-2 (KMT2A::AFDN, a non-elongation partner) is the line where menin eviction is weakest (retains 37.2% of its own top windows against 20.5% and 13.0%) and where the DOT1L track moves most of any track in the beat (retains 31.6%, 13.69 percentile points below a rank-matched null at p = 3e-10) - the opposite pattern to what an AF9/ENL-free fusion would predict, MOLM13 (KMT2A::MLLT3) is the only line in GSE127507 with a matched input in BOTH conditions, and it gives the one differential-input-controlled measurement in the series - MLL1-N occupancy at its own top 0.1% of bins is unchanged (share ratio 0.9438, input 0.9542, input-corrected 0.9891) with ZERO gained bins genome-wide, while menin retains only 13.0% of its own top windows, Q117 predicted an MLL1 ChIP would diverge from FLAG at MYB and agree at MEIS1 - both halves fire in RS4;11 (MYB 1.21 against mouse 0.71, MEIS1 0.25 against 0.23), but the MYB divergence does NOT reproduce in MOLM13, which carries the same fusion as Q117's mouse, so the MYB difference is a partner difference and not an antibody artefact, Tag the endogenous KMT2A::AFF1 allele in RS4;11 and ChIP the tag with and without VTP-50469 alongside MENIN and MLL1-N in the same chromatin - the one experiment that converts every occupancy claim in this literature from a reagent statement into a fusion statement, The deposited GSE127507 processed signal table is an Excel export in which every value at or above 1,000 is a quoted comma-separated string, so a plain numeric parse silently zeroes 34,955 of 386,298 cells - and those cells are the entire top of all eighteen distributions, Under VTP-50469 in RS4;11, the 209 bins that rise to the top of the MENIN distribution are 3.7-fold CpG-island-enriched over a class-matched null at p = 5.2e-07, while the MLL1-N gained set is not enriched at all - so menin's residue is island-biased and an MLL1-N gain is not
199. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     USP22 is a novel vulnerability regulating MEIS1 protein abundance and gene transcription in KMT2Ar acute leukemia (GSE337426)

     GEO submitters · NCBI Gene Expression Omnibus 2026
     used by An unstated system is not a coin flip: four out of five resolve away from the index disease, so silence in this graph is systematically optimistic, The unnamed KMT2A-rearranged B-ALL line is SEM, and the deposition says KMT2A::AFF1 in a genotype field, The unstated-system worklist: 26 sources carry a caveat of this class, 18 now resolve, and 13 of 16 new resolutions are imports
200. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR (Chronos) gene effect and gnomAD constraint via the Open Targets Platform and gnomAD APIs

     Broad Institute DepMap; gnomAD; Open Targets · Databases 2026
     used by Analysis (planned) — comparative druggability scan of every obligate fusion-partner interface in the atlas, EED - the PRC2 subunit this disease actually needs, with drugs in the clinic and zero leukaemia studies, ELP1 (IKBKAP/IKAP) - the Elongator scaffold that was first named for the IKK complex, is a paediatric-cancer predisposition gene, and has an RNA Pol II contact in IntAct, ELP2 (STATIP1/StIP1) - the double beta-propeller of the Elongator core, whose informal name collides with a different human gene and whose STAT3 function is a by-similarity annotation, ELP3 - the catalytic subunit, the only Elongator protein with a real acetyltransferase domain, and the one whose loss in mouse blood causes marrow failure and then p53-mutant leukaemia, ELP5 (DERP6) - the bridge that joins ELP3 to ELP4, the most essential Elongator subunit in DepMap, and a gene whose cytoplasmic call rests on an overexpression experiment, ELP6 (TMEM103/C3orf75) - an Elongator subunit named after a membrane protein it is not, whose entire localisation record comes from a Drosophila paper its own symbol cannot retrieve, JDI-12 - published as a JMJD1C inhibitor that preferentially kills KMT2A-rearranged leukaemia and spares cord blood, then used as 'KDM3i' to CREATE KMT2A rearrangements in cord-blood progenitors, by two papers that do not cite each other, JMJD1C / KDM3C - named in three entries of this graph as one of the only loci that loses fusion occupancy under a menin inhibitor, ranked 23rd of 17,728 for KMT2A-selective dependency in this project's own scan, with the strongest published shRNA effect in SEM - and no entry until now, KDM3B - the H3K9me1/2 demethylase that sits on the KMT2A breakpoint cluster region and keeps it from rearranging, and the only leukaemia gene in this graph whose direction of effect flips between the myeloid and the lymphoid literature, One demethylase family sits at both ends of this disease - KDM3B gates whether KMT2A rearranges, JMJD1C is what the resulting fusion needs to survive - and a single compound hits both, so its therapeutic index runs in opposite directions in the two cell populations it reaches, Point the EED inhibitor at the index fusion - a clinical-stage drug, a subunit with no paralogue, and zero leukaemia experiments, PRC2 - the complex, not the subunit, is the unit every experiment in this disease actually resolves, PRC2 is the fusion's repressive arm, recruited through menin, and the index disease's version of it runs to HDAC7, q172's four-line index arm is three KMT2A::AFF1 lines from TWO donors plus an ENL line; dropping the ENL line moves MEN1 7.7x more than DAXX, and menin's lost significance is variance, not effect, SMYD3 - a lysine methyltransferase with three claimed histone marks and one robust non-histone substrate, sitting on an HOXA9 feed-forward loop in MLL-AF9 AML and a fatty-acid-oxidation axis in CML stem cells, with zero interventional trials, SUZ12 - the scaffold that holds PRC2 together and decides which PRC2 it is, required in this disease and studied in it exactly once, The PRC2 dependency profile, lineage-split and expression-controlled - and the first time this project's paralogue rule came back positive
201. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     Standing instrument corrections for DepMap analyses in this project (QUEEN_BRIEF.md, cycles 41 and 45)

     Oncofusion knowledge graph coordinator · Oncofusion knowledge graph (docs/QUEEN_BRIEF.md) 2026
     used by Analysis (planned): is CLEC12A a selective dependency in KMT2A-rearranged lines in DepMap, lineage-split and reported as raw effect?, CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce, GSE159608 — 419 automated CUT&Tag libraries across 7 KMT2A-rearranged primary patient samples, 5 lines and a normal CD34+ donor, including KMT2A fusion occupancy, a menin-inhibitor arm in SEM, and three single-cell CUT&Tag libraries, KMT2A and several of its fusion partners occupy the same spatial cluster in individual nuclei before any break occurs — single-cell diploid Hi-C, RUNX1-dependent, and 3D architecture does NOT predict which fusion type a cell will make, The bounded-absence population went 39 to 73 in two days, and the abstract-level defect class went from 26 of 39 to 2 of 34 (Fisher p = 1.3e-5), The corpus citation index reads `PMID nnnnnnnn` in prose and never the `pmid` citation field — it sees 1,908 of 3,991 papers, undercounts 872 of the rest, and its headline single-cell absence was an artefact, The CSF1R-positive fetal lymphoid-primed multipotent progenitor — a fetal-restricted subset of LMPPs that makes AML rather than ALL when KMT2A::MLLT3 is switched on, and exists in human embryos, The HSPC-like blast fraction is the compartment that survives menin inhibition, and its size at diagnosis predicts depth and durability of response
202. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41351513
     Outcomes of relapsed or refractory acute myeloid leukemia after menin inhibition failure

     Multicentre retrospective consortium · Blood Advances 2026
     used by Analysis (runnable): build the first correlation matrix of menin-inhibitor resistance routes, from screens and from patient cohorts, Analysis (runnable): value a menin inhibitor as a short-duration instrument — cohort meta-regression plus a discrete-event model of remission months, Bounded absence: no menin-inhibitor trial reports response rate by 3-prime fusion partner in the primary literature, Buying an already-impaired asset — what diligence would you do, and what is the biological equivalent?, Duration and reinvestment risk — a 4.6-month response is a short-duration instrument, and short duration changes everything about how you value it, Hypothesis: a single dormant leukemia stem cell state confers cross-resistance to menin inhibition, venetoclax and T-cell engagement at once — the common-cause failure that orthogonal combination therapy assumes away, Hypothesis: only a mechanistically orthogonal partner pre-empts menin-inhibitor resistance; a second chromatin agent will not, Hypothesis: prior therapy itself installs menin-inhibitor resistance, and the dose of prior therapy predicts it better than any mutation does, Hypothesis: the survival benefit of menin inhibition is governed by the rate at which it delivers patients to a curative consolidation, not by its remission rate — and the field is measuring the wrong thing, MEN1 drug-interface mutations — the first proof that a chromatin drug can drive Darwinian escape in patients, Signalling co-mutations (RAS pathway, FLT3-ITD, WT1) under menin-inhibitor pressure, Switching menin inhibitors: four substitutions defeat all of them, three defeat only some, and the map now exists, The check this project named nine cycles ago and never ran: the 34 least-verified papers in the corpus are attached to the right claims, and every low-scoring pair is support living in a supplement or a structure rather than an abstract, TP53 inactivation and the apoptotic threshold — the drug still switches the program off, and the cell still does not die
203. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42371796
     Menin-MLL inhibitors enhance JUND activity in MLLr leukemic cells contributing to tumorigenesis and therapy resistance

     Ozyerli-Goknar E, Perner F, Armstrong SA, Zeiser R, Timmers HTM, et al. · Blood 2026
     used by Analysis (runnable): can a menin ligand tell KMT2A from JUND? A structural and co-folding selectivity profile of the whole inhibitor class, Bounded absence: no published study has tested whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor - and the same pipelines found gains for two other proteins when they looked, Bounded absence: not one menin-inhibitor study, trial or preclinical paper has ever recorded reciprocal fusion status, Census of every published chromatin-occupancy measurement made under a menin inhibitor: eight datasets, every one TSS-anchored, every one tabulating losses only, Experiment: map what stays on chromatin when the menin inhibitor has done everything it can do, Fiduciary conflict of interest — law does not remove a conflicted agent, it separates the mandates, How menin silences JunD: pocket binding plus uncoupling from MAP kinase signalling, Hypothesis: a defined subset of fusion binding sites survives saturating menin inhibition on menin-independent tethering, and that residual set is where bypass resistance comes from, Hypothesis: menin inhibitors differ in how much JUND they liberate, and the least JUND-liberating compound will be the most durable, Infants whose leukaemia expresses the reciprocal fusion survive far better - 62.4% versus 11.7% event-free at four years, JUND - the AP-1 subunit menin holds down, and the one a menin inhibitor lets go, JUND liberation — the drug's mechanism of action and one of its mechanisms of resistance are the same event, Menin inhibition is a reshuffle, not a subtraction: four proteins are documented to GAIN chromatin or persist under the drug, and none of them is the fusion, The menin–KMT2A grip (MBM1), the interface the drugs actually block, Working the queue deferred for four cycles: the 40 are three different things, and one absence claim the graph had already refuted ten entries away
204. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 41505662
     Sustained Benefit of Blinatumomab in Infants With KMT2A-Rearranged ALL: Long-Term Outcomes, Toxicity, and Pharmacokinetics

     Vieira Martins M, de Lorenzo P, Kotecha RS, et al. · Journal of Clinical Oncology 2026
     used by Analysis (runnable): fit the relapse hazard function in KMT2A-rearranged infant ALL and test whether the cohort is a mixture, Blinatumomab in infant KMT2A-rearranged ALL — the largest single outcome gain in thirty years, Hypothesis: KMT2A-rearranged infant ALL is two populations, not one — a mixture-cure model will fit better than proportional hazards, and combination therapy only needs to be given during the first year, Hypothesis: the myeloid clone that appears after CD19-directed therapy was already there at diagnosis, and its abundance is a pre-treatment biomarker, Kotecha group, Perth: the infant KMT2A-rearranged laboratory this graph had cited once, Lineage switch as antigen escape from CD19-directed therapy in KMT2A-rearranged leukemia, Nine survivors in ten carry a long-term complication, and it is worse the younger they were, Rob Pieters, Inge van der Sluis & the Interfant consortium (Princess Maxima Center, Utrecht) — the only global infant-ALL trial network, When infant leukaemia comes back, it comes back early and while treatment is still running
205. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities counted-absence searches for genome-feature intersections at the KMT2A breakpoint region, run 2026-09-20; five queries, all count 0, querytranslation inspected for each

     Q32-break-machinery · NCBI E-utilities 2026
     used by Analysis (runnable): is anything actually enriched at KMT2A breakpoints? The test has never been run against a background that could say no
206. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Ensembl REST lookup, genes ENSG00000118058 (KMT2A) and ENSG00000172493 (AFF1), assembly GRCh38, retrieved 2026-09-20

     Ensembl · Ensembl REST API 2026
     used by Analysis (runnable): is anything actually enriched at KMT2A breakpoints? The test has never been run against a background that could say no
207. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UCSC Genome Browser RepeatMasker track, GRCh38/hg38

     UCSC Genome Browser · UCSC Genome Browser 2026
     used by Analysis (runnable): is anything actually enriched at KMT2A breakpoints? The test has never been run against a background that could say no
208. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities counted-absence searches for KMT2A junction microhomology, run 2026-09-20; queries and querytranslation recorded in data.queries

     Q32-break-machinery · NCBI E-utilities 2026
     used by Analysis (runnable): mine the KMT2A junction sequences for repair-pathway signature, and count the one junction class with a genetic control behind it
209. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR (Chronos) gene effect, gnomAD constraint and Human Protein Atlas expression

     Broad Institute DepMap; gnomAD; Human Protein Atlas; Open Targets · Databases 2026
     used by Analysis: cross-validating the single-group non-menin KMT2A-fusion dependencies in one framework, MEN1 / menin — the adaptor the fusion cannot work without, MEN1 and CDK9 each carry exactly one genome-wide-significant in-gene common variant and PSIP1 two, so GWAS says almost nothing about on-target liability - yet the revumenib label lists hyperparathyroidism, the organ MEN1 loss famously affects
210. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     UniProtKB Q9BTT0 (AN32E_HUMAN), Acidic leucine-rich nuclear phosphoprotein 32 family member E, 268 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by ANP32E - the eviction chaperone: the only protein in this beat whose job is to take a histone variant OUT of chromatin
211. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     docs/memos/2026-09-22-q114-suv39h1-at-the-locus.md and ana-q114-the-four-mcf7-suv39h1-tracks-barely-agree-with-each-other-and-estradiol-changes-nothing

     Q114-suv39h1-at-the-locus · oncofusion-knowledge-graph 2026
     used by ANSWERED from the paper's own methods, not from the corresponding author: the `dsgf` token that splits GSE139196's four SUV39H1 tracks into the two high and the two low ones is DSG-plus-formaldehyde double crosslinking - so the 44.5-percentile spread Q114 reported is a protocol variable, not replicate noise
212. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/do-the-other-marks-have-two-processing-states/run.py and results.json, this repository, computed 2026-09-23 with zero downloads; channel identity for GSM8123407-412 from cycle 101 and Q125 criterion I4_rep1_channel_identity

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Answering a question three cycles filed and none ran: H3K79me2 also has two channels in GSE260742 - rep1 and rep2, 199 numeric leaves each - so cycle 104 quoted one of two available numbers without saying so. Unlike H2AK119ub they never materially disagree, 0 of 18 comparable ratios, and H3K27me3 and H3K4me3 were never scored under any second channel at all
213. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42121261
     Tissue- and tumor-type-specific expression of internal-promoter-driven YEATS-domain-devoid isoforms of MLLT1 and MLLT3.

     Dailhau S, Zamani A, Pfeifer R, et al. · Biomark Res 2026
     used by Anti-AF9 antibodies see three proteins, not two - and the third is an internal-promoter isoform discovered in 2025, Anti-ENL antibodies are the cleanest partner reagent in this disease and sit on top of the worst confound, BCOR, not CBX8, is the AHD contact MLL-AF9 cannot lose - and the same two partners bind ENL far more tightly than AF9, MLLT3 runs normal stem-cell self-renewal and is required in essentially no leukaemia line — the inverted case, The KMT2A::MLLT1 mechanism sweep: what was searched, what was read and not used, and the three retrieval traps that would have produced confident wrong answers
214. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41365398
     Fludarabine Pharmacokinetic Model-Based Dosing Minimizes the Risk of Graft Rejection in Pediatric Patients Undergoing Allogeneic Hematopoietic Cell Transplantation.

     Dvorak CC, Waters E, Temple WC, Salinas Cisneros G, Chu J, Winestone LE, Higham CS, Kharbanda S, et al. · Transplantation and Cellular Therapy 2026
     used by Anti-thymocyte globulin - a rabbit antibody against human T cells, given to infants at a fixed mg/kg dose against a T-cell mass that nobody measures, Fludarabine - the immunosuppressive half of modern conditioning, given as a phosphate prodrug, cleared by the kidney, and dosed by a body-surface-area rule that measurably fails in children, The fludarabine prodrug chain - three molecules called by one name, and the one that is measured is neither the one given nor the one that works, Weight or surface area - the Interfant-06 conditioning week uses both units in eight days, and the one measured correction in the literature went the wrong way
215. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.database
     IntAct molecular interaction database, PSICQUIC MITAB 2.7, query id:P63010 AND id:Q9UHB7, 3 rows retrieved 2026-09-21

     · IntAct (EMBL-EBI) 2026
     used by AP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology
216. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed counted-absence searches for AP2B1 against leukaemia and the KMT2A/PICALM fusion families, run 2026-09-21

     · PubMed (NCBI) 2026
     used by AP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology
217. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P63010 (AP2B1_HUMAN, AP-2 complex subunit beta)

     · UniProtKB/Swiss-Prot 2026
     used by AP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology
218. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42748909
     A dual-module ECAS complex couples transcription elongation and RNA processing

     Xiang Y, Wang H, Yin Z, et al. (Liang K, senior author) · Molecular Cell, online ahead of print 16 September 2026 2026
     used by AQR (Aquarius) - a spliceosomal RNA helicase whose loss impairs MLL-rearranged leukaemia stem cells, EAF1 — the interactor with an actual subnuclear phenotype: an MLL fusion evicts it from Cajal bodies, and the bodies themselves fall apart, ELL2 — a SEC subunit that almost no cell needs, next to a paralogue almost every cell needs, Interactome round 4: what was counted, what was written, what was rejected, and the two alias traps this beat fell into and caught, KMT2A-rearranged cells need the factors retained through elongation MORE and the factors ejected at pause release LESS — a six-gene asymmetry nobody has reported, Measure the pausing index at KMT2A::AFF1 target genes — the one number the whole pausing argument turns on, and it has never been reported, NELFE (RDBP) — the RNA-binding subunit of the pausing brake, and the alias that inflated a record count fifty-fold before it was caught, NUDT21 / CFIm25 (O43809) - the CFIm subunit that picks poly(A) sites, in the pan-MLL-fusion core, and a paraspeckle protein, SUPT5H (SPT5) — the pausing factor the KMT2A field has never studied, and the counted absence that undercuts 'the fusion hijacks pause release'
219. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O60306 (AQR_HUMAN), RNA helicase aquarius

     · UniProt Consortium 2026
     used by AQR (Aquarius) - a spliceosomal RNA helicase whose loss impairs MLL-rearranged leukaemia stem cells
220. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct REST findInteractions/P10275 (AR), 756 records, 5 against core baits; and findInteractions/22722839, 11 records; fetched 2026-09-21

     · EMBL-EBI IntAct 2026
     used by AR / androgen receptor - the best-supported partner in the absence set by independent-publication count, and the reason a menin inhibitor is in prostate cancer trials
221. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P10275 (ANDR_HUMAN) SUBCELLULAR LOCATION comment, fetched 2026-09-21

     · UniProt Consortium 2026
     used by AR / androgen receptor - the best-supported partner in the absence set by independent-publication count, and the reason a menin inhibitor is in prostate cancer trials
222. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     Human Protein Atlas v25.1, subcellular (immunofluorescence) and Blood Atlas records for AR, fetched 2026-09-21

     · Human Protein Atlas 2026
     used by AR / androgen receptor - the best-supported partner in the absence set by independent-publication count, and the reason a menin inhibitor is in prostate cancer trials
223. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q96LA8 (ANM6_HUMAN) and Q99873 (ANM1_HUMAN) FUNCTION comment blocks, both read back 2026-09-22

     · UniProt 2026
     used by Arginine methylation gates KMT2-complex access to chromatin at two residues, in both directions, independently of the menin tether AND of the WIN motif - a parallel control point, not a rival to menin dependence, H3R2me2a - the asymmetric arginine mark two residues from H3K4 that LOCKS the WRAD module off the histone tail, and the first arginine mark node in this graph, PRMT6 - the enzyme whose mark is MUTUALLY EXCLUSIVE with the MLL complex on the H3 tail, and which shuts HOX genes down by evicting WDR5
224. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed esearch counts, querytranslations and warninglists for all four H3R2/H4R3 dimethyl states against KMT2A, MLL, the fusion names, menin and revumenib, run twice 2026-09-22

     · NCBI PubMed 2026
     used by Arginine methylation gates KMT2-complex access to chromatin at two residues, in both directions, independently of the menin tether AND of the WIN motif - a parallel control point, not a rival to menin dependence
225. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O14497 (ARI1A_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by ARID1A (BAF250A) - required in the MLL-AF9 screen, present at the Myc enhancers, and flat in every selectivity measure
226. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O14497 (ARI1A_HUMAN), SUBCELLULAR LOCATION comments plus the NLS motif at 1368-1387 with evidence codes and sources, fetched live 2026-09-22

     UniProt Consortium · Databases 2026
     used by ARID1A (BAF250A) - required in the MLL-AF9 screen, present at the Myc enhancers, and flat in every selectivity measure
227. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 41661654
     ARID1A Mediates SWI/SNF-Independent Maintenance of Heterochromatin Architecture to Restrain Viral Mimicry and Immunogenicity in Colon Cancer.

     Li Q, Zhang Z, Wang Y, Peng X, Fang Y, Zhang Y, Chen L, Huang T, Yang Z, Li C, Li L, Mills GB, Shen X, Wang H, Shen J · Cancer Res 2026
     used by ARID1A (BAF250A) - required in the MLL-AF9 screen, present at the Myc enhancers, and flat in every selectivity measure
228. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q8NFD5 (ARI1B_HUMAN), AT-rich interactive domain-containing protein 1B

     UniProt Consortium · UniProt 2026
     used by ARID1B (BAF250b) - the cBAF paralogue whose deletion PROMOTES MLL-AF9 leukaemogenesis, which makes it the one gene in this beat whose published KMT2A-fusion result points the wrong way for a drug target
229. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record HGNC:18040 (ARID1B) and Ensembl gene ENSG00000049618 on GRCh37 and GRCh38

     HUGO Gene Nomenclature Committee; Ensembl · genenames.org / Ensembl 2026
     used by ARID1B (BAF250b) - the cBAF paralogue whose deletion PROMOTES MLL-AF9 leukaemogenesis, which makes it the one gene in this beat whose published KMT2A-fusion result points the wrong way for a drug target
230. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Ensembl GRCh37 REST: lookup/id/ENSG00000049618?expand=1, per-transcript start coordinates and strand

     Ensembl · Ensembl GRCh37 archive 2026
     used by ARID1B (BAF250b) - the cBAF paralogue whose deletion PROMOTES MLL-AF9 leukaemogenesis, which makes it the one gene in this beat whose published KMT2A-fusion result points the wrong way for a drug target
231. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/depmap-arid1b-paralogue-buffering/run.py and results.json, this repository; DepMap Chronos via Open Targets GraphQL, ARID1A variant annotation from Cellosaurus, pulled 2026-09-23

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by ARID1B is not dispensable - it is paralogue-buffered. In ARID1A-mutant lines its CRISPR dependency goes from -0.0724 to -0.3709 (p=0.0002) while ARID1A's own falls away (+0.1925, p=0.044) and ARID2/SMARCA4 stay flat. Cycle 80's 'dispensable everywhere' framing was wrong and the caveat that entry carried was right
232. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/depmap-arid1b-leave-one-out-and-lineage-concentration/run.py and results.json, this repository; DepMap Chronos via Open Targets GraphQL, ARID1A variant annotation from Cellosaurus, pulled 2026-09-23

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by ARID1B's buffering effect survives leave-one-out in 11 of 11 - even dropping the line whose own effect is -1.2269 leaves -0.2129 at p = 0.00465. But the lineage-concentration test says there was LITTLE confound to remove, cycle 84's headline confirms an absence rather than defeating a threat, and my own pre-registered TVD threshold fired on a measure that is large by construction at n=11
233. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q68CP9 (ARID2_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by ARID2 (BAF200) - the PBAF subunit whose loss ACCELERATES MLL-AF9 leukaemia, and which is nonetheless required to maintain it
234. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.database
     DepMap KMT2A-selective dependency scan over 17,728 genes: analyses/depmap-kmt2a-selective-scan-full/results.json, generated 2026-09-19. ARID4B rank 11,283, difference +0.0164, mean_kmt2a_r -0.1472, mean_other -0.1635, pan_essentiality -0.1548.

     · DepMap via this repository 2026
     used by ARID4B / SAP180 (Q4LE39) - the Sin3A-complex subunit that shares SIN3A's HDAC1-binding region, is in the MLL-fusion interactome, and is the second of only two epigenetic vulnerabilities a CRISPR screen found in KMT2A-rearranged ALL
235. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     NCBI esearch via scripts/pubmed_safe.py on 2026-09-24. ARID4B[tiab] AND MLL[tiab] -> 0 (translation "ARID4B"[Title/Abstract] AND "MLL"[Title/Abstract]). ARID4B[tiab] AND KMT2A[tiab] -> 1 (37686014). ARID4B[tiab] AND leukemia[tiab] -> 3 (37686014, 26189108, 18728284). ARID4B[tiab] AND SIN3A[tiab] -> 2 (33907498, 33454011). ARID4B[tiab] AND NuRD[tiab] -> 0. SAP180[tiab] -> 2 (26609676, 12724404). BRCAA1[tiab] AND leukemia[tiab] -> 0. Positive control MEN1[tiab] AND MLL[tiab] -> 55.

     · NCBI PubMed E-utilities 2026
     used by ARID4B / SAP180 (Q4LE39) - the Sin3A-complex subunit that shares SIN3A's HDAC1-binding region, is in the MLL-fusion interactome, and is the second of only two epigenetic vulnerabilities a CRISPR screen found in KMT2A-rearranged ALL
236. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42100853
     ASF1B promotes erythropoiesis by regulating the establishment and enrichment of H3.3 nucleosomes.

     Liu J, Song X, Zhou L, et al. · Nucleic Acids Research 2026
     used by ASF1A - the compensating paralogue: the only gene in this beat that is not required in a single one of 1,208 cell lines, and it still has a leukaemia phenotype, ASF1B - the paralogue that actually does the work in blood: it occupies >70% of H3.3 nucleosomes in fetal liver while HIRA is dispensable there, HIRA - the chaperone that puts new H3.3 into transcribed gene bodies, and therefore the machine that runs this graph's clearance clock
237. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     Is any histone chaperone or variant-exchange gene selectively required in KMT2A-rearranged leukaemia lines, and does the answer survive a lineage split?

     · Oncofusion knowledge graph (internal analysis) 2026
     used by ASF1A - the compensating paralogue: the only gene in this beat that is not required in a single one of 1,208 cell lines, and it still has a leukaemia phenotype, ASF1B - the paralogue that actually does the work in blood: it occupies >70% of H3.3 nucleosomes in fetal liver while HIRA is dispensable there, ATRX - DAXX's partner and the other arm of H3.3 deposition; in this disease its only signal is in the four lymphoid cell lines, CABIN1 - a 2,220-residue calcineurin inhibitor that is also a full member of the H3.3 chaperone complex, and nobody has asked which job matters in blood, CHAF1A (CAF-1 p150) - the subunit that reads PCNA and makes histone deposition replication-coupled, and the half of CAF-1 that leukaemia has not looked at, CHAF1B (CAF-1 p60) - the one histone chaperone in this beat with a real leukaemia literature, and a mouse experiment that contradicts DepMap's verdict on its druggability, H2AZ1 - the one histone variant with a direct KMT2A-fusion paper: MLL-AF10 and MLL-ENL recruit Tip60 to Hoxa9 to acetylate H2A.Z, HIRA - the chaperone that puts new H3.3 into transcribed gene bodies, and therefore the machine that runs this graph's clearance clock, Is any histone chaperone selectively required in KMT2A-rearranged leukaemia? Twenty-nine genes, lineage-split - and the answer is no, with one number worth a null and a positive tail nobody expected, NASP - the soluble histone reservoir, and the only gene in this beat with a real selective difference at a pan-essentiality shallow enough to interpret, SUPT6H (SPT6) - the histone chaperone that rides with RNA polymerase II, which puts it closer to the fusion's proposed mechanism than anything else in this beat, The HUCA complex (HIRA/UBN1/CABIN1/ASF1a) - four proteins, one job: put H3.3 into chromatin without waiting for S phase, UBN1 - the subunit that decides the HIRA complex is carrying H3.3 and not H3.1, by reading a single glycine
238. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9Y294 (ASF1A_HUMAN), Histone chaperone ASF1A, 204 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by ASF1A - the compensating paralogue: the only gene in this beat that is not required in a single one of 1,208 cell lines, and it still has a leukaemia phenotype
239. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9NVP2 (ASF1B_HUMAN), Histone chaperone ASF1B, 202 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by ASF1B - the paralogue that actually does the work in blood: it occupies >70% of H3.3 nucleosomes in fetal liver while HIRA is dispensable there
240. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9UBL3 (ASH2L_HUMAN), retrieved via the UniProt REST API

     UniProt Consortium · UniProtKB 2026
     used by ASH2L — the WRAD subunit that has DPY30's essentiality profile and none of its co-dependency signal
241. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 3TOJ and 4RIQ, verified against the RCSB data API for title, experimental method and resolution

     RCSB Protein Data Bank · RCSB PDB 2026
     used by ASH2L — the WRAD subunit that has DPY30's essentiality profile and none of its co-dependency signal, DPY30 — the smallest COMPASS subunit, and the only one that behaves like a selective dependency, Separate positioning from catalysis: an ASH2L mutant that keeps RBBP5 and loses DPY30
242. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     dependency-claims-vs-own-scan/run.py: every disease-scoped requirement claim checked against the project's own genome-wide DepMap selectivity scan

     · Oncofusion knowledge graph (internal analysis) 2026
     used by ASH2L and RBBP5 rank 17,248 and 14,622 of 17,728 for KMT2A-selectivity, which is what this graph's own SET-domain-is-dispensable thread predicts - the audit found agreement where it looked for contradiction
243. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     subtraction-kmt2a-set-domain: the SET domain (aa 3829-3945), the lost module whose loss demonstrably does not matter

     · Oncofusion knowledge graph (internal entry) 2026
     used by ASH2L and RBBP5 rank 17,248 and 14,622 of 17,728 for KMT2A-selectivity, which is what this graph's own SET-domain-is-dispensable thread predicts - the audit found agreement where it looked for contradiction
244. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     UCSC Genome Browser API, mm10 refGene track, per-interval query for each of the 52 bins, run 2026-09-23

     · UCSC Genome Browser 2026
     used by Asked for the first time whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor: all three pre-registered criteria fire, and the set they fire on is 52 low-amplitude bins reaching 6x the genome median, 16 of them with no gene within 20 kb and 1 of 52 on a CpG island where 66% of the fusion's real sites are
245. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UCSC Genome Browser API, mm10 cpgIslandExt track (15,997 islands on chr1-19 and chrX), joined to GSE260742 FLAG-MLLAF9 2 kb bins, run 2026-09-23

     · UCSC Genome Browser 2026
     used by Asked for the first time whether a KMT2A fusion GAINS chromatin sites under a menin inhibitor: all three pre-registered criteria fire, and the set they fire on is 52 low-amplitude bins reaching 6x the genome median, 16 of them with no gene within 20 kb and 1 of 52 on a CpG island where 66% of the fusion's real sites are, Menin inhibition removes the fusion from its strongest sites without relocating it to a new programme, and PRC1.1 fills in only where the eviction is complete - so the drug's chromatin action is subtraction plus local substitution, not redistribution, The falsifier that decides the beat, run: 65.6% of the fusion's own top sites overlap a CpG island and 1 of the 52 sites it formally 'gains' under a menin inhibitor does - 1.35-fold over a class-matched null, Fisher p = 0.53 - so the gained set does not carry the signature of CXXC-driven binding and the redistribution model is refuted on its own predicted substrate
246. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42241118
     Direct pharmacological targeting of asparagine synthetase to overcome resistance to L-asparaginase in ALL therapy

     Claude R, ..., (see record) · JCI Insight 2026
     used by ASNS - the resistance gene for a backbone drug, never measured in this disease
247. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human record P08243 (ASNS, asparagine synthetase)

     · UniProt 2026
     used by ASNS - the resistance gene for a backbone drug, never measured in this disease
248. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     The B-lineage ALL comparator arm enumerated with Cellosaurus accessions, fusions and DAXX Chronos values

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Assay telomere maintenance across the thirteen B-lineage ALL DepMap lines, and see whether it predicts the DAXX dependency split
249. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     Q154 genome-wide gene-body occupancy matrix and label permutations

     Iqbal Lab oncofusion knowledge graph, beat Q154 · this project 2026
     used by At gene-body scale the AFF1-exclusive recurrent class is 6 against an exact institution-stratified null of 0.86 - but the singleton class is 63 against 74, so recurrence carries the signal and exclusivity alone does not, IgG passes this project's occupancy gate at TWICE the rate of the fusion antibody - because the gate is normalised to each library's own promoter mean and IgG has almost none, Run on the reciprocal KMT2A-C antibody instead of the fusion-side KMT2A-N, the same rule gives zero AFF1-exclusive recurrent gene bodies against a null of 0.52 - the gene-scale class is fusion-arm specific, The same occupancy rule passes 0.61% of gene bodies and 10.75% of promoters, so IRX1 at 3 of 27 is 18x its gene-scale expectation where at promoter scale it was below it, Verdict: q147's IRX1 headline survives as a measurement and not as an inference - the pre-registered branch is NEITHER, on U5 and on a FALLS condition that fired
250. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - P18847, ATF3_HUMAN, cyclic AMP-dependent transcription factor ATF-3

     · UniProt Knowledgebase 2026
     used by ATF3 (P18847) - the one cascade member the DOT1L inhibitor hits HARDER than the menin inhibitor, and a transcriptional REPRESSOR being counted as a stress output
251. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42350371
     Mitochondrial integrated stress response activation creates a therapeutic vulnerability to MCL-1 inhibition in acute myeloid leukemia.

     Brakefield-Laird L, Budhraja A, Hall PM, Brewington D, et al. · Cell Death & Disease 2026
     used by ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor, Complex I inhibition triggers the mitochondrial integrated stress response through DELE1 and HRI, and that is what creates the MCL-1 dependency - a metabolic drug manufacturing an apoptotic vulnerability, MCL1 - named in 34 entries of this graph with no entry of its own until now; a mitochondrial survival protein high in infant KMT2A-rearranged ALL, druggable, and so far undruggable in patients, Mitochondrial complex I plus venetoclax in KMT2A-rearranged paediatric AML - a genotype-stratified in vivo PDX result, on a drug whose adult clinical development stopped for toxicity
252. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB - Q9NZJ5, E2AK3_HUMAN, eukaryotic translation initiation factor 2-alpha kinase 3 / PERK

     · UniProt Knowledgebase 2026
     used by ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor, EIF2AK3 / PERK (Q9NZJ5) - the kinase whose transcript is flat under both inhibitors, whose arm IS functionally required in MLL-AF9 leukaemia stem cells, and whose common name collides with phospho-ERK
253. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - P18848, ATF4_HUMAN, cyclic AMP-dependent transcription factor ATF-4

     · UniProt Knowledgebase 2026
     used by ATF4 / CREB2 (P18848) - the hub this graph inferred from its targets and never wrote, and its own transcript does not move under either menin inhibitor
254. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - Q9Y2D1, ATF5_HUMAN, cyclic AMP-dependent transcription factor ATF-5

     · UniProt Knowledgebase 2026
     used by ATF5 / ATFX (Q9Y2D1) - the most reproducible repression in the earlier dataset, and the three probe values the graph quotes are SEM, not RS4;11
255. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     hein-abundance-versus-confidence: does prey abundance predict low confidence class specifically, or every class equally?

     · This repository (reproducible from the Hein 2015 supplementary workbooks) 2026
     used by Atlas confidence class tracks the prey:bait ratio, not abundance - so the proposed copy-number triage of the atlas-only partners does not exist, Atlas-only interactome partners are predicted by how abundant the prey is, not by which bait was pulled - and if so this graph's interactome denominator is inflated
256. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q13315 (ATM_HUMAN), Serine-protein kinase ATM, 3056 aa

     · UniProt Consortium 2026
     used by ATM - the DNA-damage kinase that phosphorylates three of the four KMT2A 3' partners
257. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Proteins API proteomics-ptm P55199: 12 phosphosites at 126, 180, 308, 309, 317, 319, 401, 437, 442, 444, 519, 561

     · EMBL-EBI 2026
     used by ATM phosphorylates ELL at T551, S561 and S589 - inside the OCEL domain this graph already calls essential for MLL-ELL immortalisation, ELL S309 is the most frequently observed phosphosite across public datasets - and it is a proline-directed site with no kinase, not an ATM site
258. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB canonical FASTA Q03111 and P42568, scanned for [ST]Q motifs and compared over residues 380-410, 2026-09-19

     · UniProt Consortium 2026
     used by ATM phosphorylates ENL at conserved SQ sites and flips it from elongation to Polycomb repression - on sequence every KMT2A::MLLT1 fusion keeps
259. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB search gene_exact:ATR AND organism_id:9606 AND reviewed:true, fields accession/id/gene_primary/gene_synonym/length/protein_name, run 2026-09-20: 3 rows - P20848 SERPINA2 421 aa, Q9H6X2 ANTXR1 564 aa, Q13535 ATR 2644 aa

     · UniProt Consortium 2026
     used by ATR - the replication-stress kinase with the best preclinical case in this disease and the best published reason not to inhibit it
260. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P46100 (ATRX_HUMAN), Chromatin remodeler ATRX, 2492 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by ATRX - DAXX's partner and the other arm of H3.3 deposition; in this disease its only signal is in the four lymphoid cell lines
261. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI Gene esummary for GeneID 153572 (IRX2) and UCSC hg38 refGene, cross-checked against C:\Temp\q145irx1\promoters.tsv

     NCBI Gene / UCSC Genome Browser / Iqbal Lab oncofusion knowledge graph beat Q153 · NCBI Gene 2026
     used by Attribution is asymmetric: q151's IRX2 negative survives a window that is 89.2% someone else's gene, and a positive in the same window would not have
262. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     IRX2 is not fusion-occupied at promoter scale in GSE252378: 0 of 27 samples, against a null of 2.73

     Iqbal Lab oncofusion knowledge graph, beat Q151 · this project 2026
     used by Attribution is asymmetric: q151's IRX2 negative survives a window that is 89.2% someone else's gene, and a positive in the same window would not have
263. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     q153 two-window deconvolution at the IRX2 / IRX2-DT locus with a shared-pair null, GSE252378

     Iqbal Lab oncofusion knowledge graph, beat Q153 · this project 2026
     used by Attribution is asymmetric: q151's IRX2 negative survives a window that is 89.2% someone else's gene, and a positive in the same window would not have, Tile chr5:2,750,677-2,752,892 at 100 bp from the GSE252378 bigWigs and ask which TSS the p454 signal sits on, Two-window deconvolution: at a typical shared promoter only 1.8% of the signal can be proved to belong to the gene it is filed under, and at IRX2's one near-positive it is 19.7%
264. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q14CW9 (AT7L3_HUMAN), SUBCELLULAR LOCATION comment with evidence codes and sources plus the feature table, fetched live 2026-09-22

     UniProt Consortium · Databases 2026
     used by ATXN7L3 — the shared adaptor that actually sets H2Bub levels, and a KMT2A-selectivity hit nobody in this field has looked at
265. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST service: article and full-text-XML availability check, run 2026-09-19

     Q7-clinical-infant beat (this graph) · Europe PMC 2026
     used by Audit of the infant presenting-feature evidence: what exists, what does not, what was rejected, what could not be read, Papers we know exist, know matter, and cannot read, Treating the five menin inhibitors as one drug, when a 2026 title says they are not
266. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities search and ELink related-articles retrieval for PMID 18676843, performed for this beat

     Q-subtraction-kmt2a; NCBI PubMed · PubMed (NCBI) 2026
     used by Audit, CORRECTED: the PHD add-back result was replicated in 2008 by a second group in a different fusion, and this graph already cited it
267. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/which-entries-rest-on-one-h2ak119ub-convention/run.py and results.json, this repository; corpus read from git HEAD 2026-09-23; convention values from the committed results of analyses/q125-the-h2ak119ub-arm-of-the-pcgf1-knockout and from cycle 104

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Auditing all 2,476 committed entries for single-convention H2AK119ub: 122 mention the mark, 10 carry both conventions, 2 name the one they use, and 2 quoted a paired-convention number without saying so. Both are amended - and the more useful result is that the two conventions disagree PER STATISTIC rather than globally, so declaring the mark unusable would have been wrong
268. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     A Groupwide Pilot Study to Test the Tolerability and Biologic Activity of the Addition of Azacitidine (IND# 133688, NSC# 102816) to Chemotherapy in Infants With Acute Lymphoblastic Leukemia (ALL) and KMT2A(MLL) Gene Rearrangement

     Children's Oncology Group · ClinicalTrials.gov 2026
     used by Azacitidine - the one hypomethylating agent actually given to infants with KMT2A-rearranged ALL: safe, demonstrably demethylating, and it did not change survival, COG AALL15P1 — azacitidine epigenetic priming in infant KMT2A-rearranged ALL: on-target, well tolerated, no survival effect
269. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42272166
     All-Oral Combination of Revumenib, Decitabine, and Venetoclax for Relapsed or Refractory AML (SAVE).

     Issa GC, Cuglievan B, El Hajjar G, et al. · J Clin Oncol 2026
     used by Azacitidine - the one hypomethylating agent actually given to infants with KMT2A-rearranged ALL: safe, demonstrably demethylating, and it did not change survival, Decitabine - degrades DNMT1, and in ALL cells it degrades KMT2A too: the hypomethylating agent with two mechanisms and a menin-inhibitor combination already in the clinic, Evicting the fusion is not enough: PRC1.1 has to arrive, and losing it causes menin-inhibitor resistance in vivo, SAVE (NCT05360160): the first prospective menin-inhibitor trial to report a NUP98-rearranged subgroup outcome
270. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ChEMBL molecule record CHEMBL1489 (AZACITIDINE)

     ChEMBL, EMBL-EBI · ChEMBL database 2026
     used by Azacitidine - the one hypomethylating agent actually given to infants with KMT2A-rearranged ALL: safe, demonstrably demethylating, and it did not change survival
271. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     A Phase 1 Study of Revumenib, Azacitidine, and Venetoclax in Pediatric and Young Adult Patients With Refractory or Relapsed Acute Myeloid Leukemia

     St. Jude Children's Research Hospital · ClinicalTrials.gov 2026
     used by Azacitidine - the one hypomethylating agent actually given to infants with KMT2A-rearranged ALL: safe, demonstrably demethylating, and it did not change survival
272. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human entry P61769 (B2MG_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by B2M (beta-2 microglobulin) - the invariant light chain every MHC class I molecule needs, the classic class I escape route, and a gene this disease's literature has never once named
273. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P61769 (B2M) SUBCELLULAR LOCATION comment, fetched 2026-09-21

     · UniProt Consortium 2026
     used by B2M (beta-2 microglobulin) - the invariant light chain every MHC class I molecule needs, the classic class I escape route, and a gene this disease's literature has never once named
274. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, 2026-09-21. querytranslation read back on every run; cytogenetic terms quoted so they expand to phrases and not to a UID filter

     · NCBI PubMed 2026
     used by B2M (beta-2 microglobulin) - the invariant light chain every MHC class I molecule needs, the classic class I escape route, and a gene this disease's literature has never once named
275. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 41734382
     Overcoming menin inhibitor resistance in AML cells with combinations including BET proteins and a dual BRG1/BRM inhibitor.

     Fiskus W, Mill CP, Issa GC, Piel J, et al. · Blood 2026
     used by BAF does not work where the fusion works: most direct MLL-AF9 targets do not respond to Brg1 loss, and the SWI/SNF address is a distal MYC enhancer cluster 1.7 Mb away, BAF inhibition against the menin axis: on the evidence it is a complement, not an alternative - and the case rests on one laboratory, FHD-286 - the dual BRG1/BRM inhibitor that carries every SWI/SNF therapeutic claim in this disease, with a minimum enrolment age of 16, Separate ncBAF from canonical BAF in KMT2A-rearranged cells: paired degron, paralogue-pair CRISPR, and complex-resolved CUT&RUN, The menin-inhibitor-resistant state has its own address: SMARCA4, BRD4 and CREBBP, with no MEN1 mutation anywhere
276. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct REST findInteractions/P46379 (BAG6), 910 records, 2 against a core bait, fetched 2026-09-21

     · EMBL-EBI IntAct 2026
     used by BAG6 / BAT3 - a DOT1L partner that is about H3K79me2 itself, reciprocally co-immunoprecipitated, and not a spoke
277. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed esearch: (DOT1L OR "Dot1L")[tiab] AND (BAG6 OR BAT3)[tiab] = 1; (BAG6 OR BAT3)[tiab] AND (MLL OR KMT2A OR leukemia)[tiab] = 8; run 2026-09-21

     · NCBI PubMed 2026
     used by BAG6 / BAT3 - a DOT1L partner that is about H3K79me2 itself, reciprocally co-immunoprecipitated, and not a spoke
278. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P46379 (BAG6_HUMAN) SUBCELLULAR LOCATION comments, note, feature table and TISSUE SPECIFICITY, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by BAG6 / BAT3 - a DOT1L partner that is about H3K79me2 itself, reciprocally co-immunoprecipitated, and not a spoke
279. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9P281 (BAHC1_HUMAN), BAH and coiled-coil domain-containing protein 1

     UniProt Consortium · UniProt 2026
     used by BAHCC1 - a BAH-domain H3K27me3 reader that an MLL fusion binds and upregulates in mouse, and the locus carrying the two largest MLL1-N-reactive rises under a menin inhibitor in human RS4;11
280. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record HGNC:29279 (BAHCC1) and Ensembl gene ENSG00000266074

     HUGO Gene Nomenclature Committee; Ensembl · genenames.org / Ensembl 2026
     used by BAHCC1 - a BAH-domain H3K27me3 reader that an MLL fusion binds and upregulates in mouse, and the locus carrying the two largest MLL1-N-reactive rises under a menin inhibitor in human RS4;11
281. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Ensembl GRCh37 REST: overlap/region/human/17:79300000-79450000?feature=gene, and lookup/id/ENSG00000266074

     Ensembl · Ensembl GRCh37 archive 2026
     used by BAHCC1 - a BAH-domain H3K27me3 reader that an MLL fusion binds and upregulates in mouse, and the locus carrying the two largest MLL1-N-reactive rises under a menin inhibitor in human RS4;11
282. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q99728 (BARD1_HUMAN) - BRCA1-associated RING domain protein 1, Homo sapiens, with sequence

     UniProt Consortium · UniProt Knowledgebase 2026
     used by BARD1 (Q99728, 777 aa) - BRCA1's obligate partner, and the member of the cascade that actually touches the chromatin: its PxVxL motif at 569-573 binds HP1, and CTCF binds BARD1 rather than BRCA1 in an undamaged cell
283. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42475683
     CTCF primes repressive chromatin for rapid BRCA1/BARD1 recruitment to drive homologous recombination repair of DNA double-strand breaks.

     Hwang SY, Choi D, Choi MJ, Jo HR, Jeong JH, Kang MA, Lee JS · Nucleic Acids Research 2026
     used by BARD1 (Q99728, 777 aa) - BRCA1's obligate partner, and the member of the cascade that actually touches the chromatin: its PxVxL motif at 569-573 binds HP1, and CTCF binds BARD1 rather than BRCA1 in an undamaged cell, BRCA1 (P38398, 1,863 aa) - named in 19 entries of this graph and never written: the HR mediator that CTCF and the H3K9 machinery are now reported to license at a break, and the only cascade member with a human therapy-related-leukaemia signal after topoisomerase II therapy, CBX3 / HP1-gamma (Q13185, 183 aa) - the EUCHROMATIC HP1 paralogue, and the one CTCF actually prefers: the reader that carries BRCA1/BARD1 onto H3K9-methylated break chromatin, flat in DepMap, and a fusion partner in one KMT2A::AFDN AML patient, DepMap across the whole CTCF-BRCA1 licensing cascade in one call: CTCF is a near-total pan-essential at -1.83, BARD1 is deeper than BRCA1, the HP1 paralogue that matters is flat while the one that does not is four times deeper - and DNA2's row has 80 screens where everything else has 1,258, DNA2 (P51530, 1,060 aa) - EXO1's redundant partner in long-range resection, dual mitochondrial and nuclear, and the gene whose DepMap row is measured in 80 screens where every other cascade member has 1,258, EXO1 (Q9UQ84, 846 aa) - the long-range resection nuclease at the bottom of the CTCF cascade, and the point where this graph's two independent 2026 HR papers meet: CtIP starts the resection, EXO1/DNA2 finish it, Giving a G9a/GLP inhibitor alongside etoposide or doxorubicin removes a DNA-repair licence at the moment TOP2 breaks are being made - and the sibling compound UNC-0638 is measured to block BRCA1/BARD1 arrival at a break, in children on curative therapy, HP1-gamma holds a constitutive, sub-threshold foothold on the KMT2A exon-11 CTCF anchor in undamaged cells - the CTCF-primed model's missing resting state - and the reason nobody has seen it is that it sits below the reproducibility filter every public peak set applies, KDM5A / JARID1A / RBP2 (P29375, 1,690 aa) - already in this graph as the 3' half of NUP98::KDM5A and never written as a gene: the H3K4 demethylase CTCF recruits to a break, the only ERASER in the licensing cascade, and the one cascade member with no blood-lineage ChIP-seq anywhere, No BRCA1, BARD1, EXO1, DNA2, HP1-gamma or KDM5A has ever been localised at the KMT2A locus, in any cell type: 21 queries, every querytranslation read back, 0 on-topic records - and the absence splits THREE ways by class, which is what decides what it costs to close, SETDB1 (Q15047, KMT1E/ESET) - the second H3K9me3 writer: it represses the fusion's own targets via the PAF1 complex, it is required in fetal liver HSPCs and in leukaemic stem cells, and it is a broad dependency in half the CRISPR panel - three published directions that do not agree, SUV39H1 (O43463, KMT1A) - the H3K9me3 writer: a menin effector whose binding region OVERLAPS the drug pocket at its N-terminus, a tumour suppressor in MLL-AF9 AML, recruited to double-strand breaks by CTCF, and never localised at the KMT2A locus in any cell type, The CTCF licensing cascade: CTCF recruits HP1-gamma, KAP1, SUV39H1, SETDB1 and KDM5A to a break, that repressive chromatin licenses EARLY BRCA1/BARD1, and early BRCA1/BARD1 licenses EXO1/DNA2 long-range resection - measured entirely in U2OS and HEK-293T, with engineered breaks and one etoposide arm, The H3K9me3 writer reaches the KMT2A anchor only through CTCF and only after the break - which would make this graph's CTCF arm and its H3K9 arm one arm, with CTCF upstream, and would make every resting-state H3K9me3 measurement at that element the wrong assay, The menin arm and the CTCF arm are not two stories: menin/SETD2/H3K36me3/LEDGF licenses CtIP short-range resection and CTCF/H3K9me/HP1-gamma licenses BRCA1/BARD1-EXO1/DNA2 long-range resection, which are the two consecutive halves of ONE reaction - so a menin inhibitor and an H3K9-axis inhibitor are acting on the same pathway at different steps
284. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ENCODE Portal experiment search, type=Experiment, assay_title=TF ChIP-seq, status=released, per target label

     ENCODE Project Consortium · ENCODE Portal 2026
     used by BARD1 (Q99728, 777 aa) - BRCA1's obligate partner, and the member of the cascade that actually touches the chromatin: its PxVxL motif at 569-573 binds HP1, and CTCF binds BARD1 rather than BRCA1 in an undamaged cell
285. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.databasePMID 42051034
     ChIP-Atlas 2025 update: 10-year anniversary of a data-mining platform for exploring epigenomic landscape.

     Zou Z, Ohta T, Kasukawa T, et al. · Nucleic Acids Research 2026
     used by BARD1 and EXO1 are the only two proteins in the resection layer with ZERO chromatin-occupancy datasets anywhere - ENCODE, ChIP-Atlas, ReMap and a hand-checked GEO title list all return nothing - and BARD1's silence is the loud one, because its obligate partner BRCA1 has thirty-one, BRCA1 at the KMT2A exon-11 CTCF anchor is real, cell-type-restricted, and survives the summit-matched null Q35 asked for: 93.7th and 92.3rd percentile of same-cell CTCF anchors in HeLa-S3 and H1, rising to 95.2 and 95.7 under width-and-TSS matching - and 34.0 and 45.1 in the only two blood lines, where ENCODE's 300,000-peak pre-IDR lists are also empty, Outside ENCODE the resection layer is not one absence but four: 31 BRCA1, 45 CBX3, 22 KDM5A, 14 MRE11 and 8 CtIP ChIP-seq experiments exist in ChIP-Atlas, DNA2 and RAD50 have one each, and only BARD1 and EXO1 have none in any of the three archives, Read all 76 archived CBX3 and BRCA1 tracks at the KMT2A exon-11 anchor at SIGNAL level against a CTCF-anchor-matched null, and score the anchor's rank among its own class in each - no wet work, no new data, and it decides whether the sub-IDR peak is a foothold or a threshold artefact, The KMT2A exon-11 CTCF anchor carries a CBX3 peak and a BRCA1 peak in three of the very ENCODE experiments Q35 called empty - the peaks are in ENCODE's own pre-IDR files, present in both replicates, and deleted by the IDR threshold at global IDR 0.227 and 0.059
286. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch/esummary db=gds, <SYMBOL>[Title] AND "Homo sapiens"[Organism] AND gsm[ETYP], querytranslation read back for every query; sample-level SOFT records fetched for the ambiguous titles

     National Center for Biotechnology Information · NCBI Gene Expression Omnibus 2026
     used by BARD1 and EXO1 are the only two proteins in the resection layer with ZERO chromatin-occupancy datasets anywhere - ENCODE, ChIP-Atlas, ReMap and a hand-checked GEO title list all return nothing - and BARD1's silence is the loud one, because its obligate partner BRCA1 has thirty-one
287. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     analyses/q37-mrn-depmap-adjudication/results.json and correlations.json (Open Targets 26.06) with analyses/q37-archive-sweep-at-the-kmt2a-locus/census.json

     Q37-archive-sweep · this graph 2026
     used by BARD1 and EXO1 are the only two proteins in the resection layer with ZERO chromatin-occupancy datasets anywhere - ENCODE, ChIP-Atlas, ReMap and a hand-checked GEO title list all return nothing - and BARD1's silence is the loud one, because its obligate partner BRCA1 has thirty-one, RAD50 (Q92878, 1312 aa) - the third MRN subunit, and the one that settles whether the complex is incoherent in DepMap: mean -0.114 with 5.2% of 1,257 lines below -0.5, so it sits with NBN and not with MRE11, yet it covaries with MRE11 across cell lines as tightly as BRCA1 does with BARD1, The MRN asymmetry adjudicated: RAD50 -0.114 sits with NBN -0.047 and not with MRE11 -0.542, so it is not one subunit against two - and yet MRE11 and RAD50 covary across 1,208 cell lines at r = 0.554, the same as BRCA1 against BARD1, against a background of zero
288. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     Selectivity against normal-tissue exposure: crossing the 17,728-gene KMT2A-rearranged dependency scan with GTEx

     · Oncofusion knowledge graph (internal analysis) 2026
     used by BCL11A — the sickle-cell target whose binding sites lose methylation in this disease, KAT6A / MOZ — the acetyltransferase that is a fusion oncogene in its own right, Selective in the disease and quiet everywhere else: MYB comes out cleanest, TERT — telomerase, switched on by the index fusion through HOXA7
289. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 2XA0, 4LVT, 6O0K and 6QGG, each verified against the RCSB data API for title, experimental method and resolution

     RCSB Protein Data Bank · RCSB PDB 2026
     used by BCL2 — the one co-dependency in this beat that already has an approved drug pointed at it
290. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P10415 (BCL2_HUMAN), apoptosis regulator Bcl-2, retrieved via the UniProt REST API

     UniProt Consortium · UniProtKB 2026
     used by BCL2 — the one co-dependency in this beat that already has an approved drug pointed at it
291. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 3BIM, 6B7G, 2N1L, 4HPL and 8HCU, with per-entity UniProt accessions and modelled residue ranges resolved against the RCSB search and data APIs

     RCSB Protein Data Bank · RCSB PDB 2026
     used by BCOR — the other side of the CBX8 conflict, and a target you would be inhibiting in the direction of disease
292. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, protein-protein-interaction target CHEMBL4106125 (BCoR-BCL6): target_search, get_bioactivity and get_mechanism queries, and the absence of any single-protein BCOR target

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by BCOR — the other side of the CBX8 conflict, and a target you would be inhibiting in the direction of disease
293. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42273916
     Sustained PRC1.1 activity preserves gene repression independently of PRC2.2 and restrains leukemic cell differentiation.

     Mojallali F; De Meis A; Alkema SG; Atsma TJ; Hogeling SM; Kloosterman R; Ioannou E; Wierenga ATJ; Huls G; Martens JHA; Schuringa JJ; van den Boom V · Nucleic Acids Res 2026
     used by BCOR — the other side of the CBX8 conflict, and a target you would be inhibiting in the direction of disease
294. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch record count for (MLL OR KMT2A) AND (BCORL1 OR CXXC10), run for this beat

     · NCBI PubMed, queried 2026-09-19 2026
     used by BCORL1 — the BCOR paralogue that is in the only crystal structure of the PRC1.1 core, and is the one PRC1.1 subunit nobody has tested in KMT2A-rearranged leukaemia
295. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P11274 (BCR_HUMAN): reviewed sequence, length and positional feature annotations

     UniProt Consortium · UniProtKB/Swiss-Prot 2026
     used by BCR — domain architecture and residue coordinates
296. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro/Pfam/CATH-Gene3D/SUPERFAMILY domain matches for UniProt P11274

     InterPro Consortium; EMBL-EBI · InterPro 2026
     used by BCR — domain architecture and residue coordinates
297. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Ensembl transcript ENST00000305877 (BCR): exon structure and coding-sequence coordinates, used to convert exon breakpoints into residue numbers

     Ensembl; EMBL-EBI · Ensembl REST API 2026
     used by BCR — domain architecture and residue coordinates
298. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     AlphaFold Protein Structure Database entry for P11274 (model AF-P11274-F1)

     Jumper J, Varadi M, et al.; EMBL-EBI/DeepMind · AlphaFold DB 2026
     used by BCR — domain architecture and residue coordinates
299. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR (Chronos) gene effect, retrieved via the Open Targets Platform GraphQL API

     Broad Institute DepMap; Open Targets · Database 2026
     used by BET bromodomain inhibition in KMT2A-rearranged leukemia — spectacular preclinically, stalled clinically
300. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9NZM4 (BICRA_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by BICRA (GLTSCR1) - the ncBAF scaffold, zero PubMed records in KMT2A leukaemia, and a lineage-independent dependency in KMT2A-rearranged lines
301. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9NZM4 (BICRA_HUMAN) SUBCELLULAR LOCATION and TISSUE SPECIFICITY comments, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by BICRA (GLTSCR1) - the ncBAF scaffold, zero PubMed records in KMT2A leukaemia, and a lineage-independent dependency in KMT2A-rearranged lines
302. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     ClinicalTrials.gov registry records for NCT05153330 (A Phase 1 First-in-human Dose-escalation and Dose-expansion Study of BMF-219, an Oral Covalent Menin Inhibitor, in Adult Patients With Acute Leukemia, DLBCL, Multiple Myeloma and CLL/SLL) and NCT05631574, retrieved via the ClinicalTrials.gov API v2

     Biomea Fusion Inc. · ClinicalTrials.gov 2026
     used by Biomea Fusion — icovamenib (BMF-219): the covalent menin inhibitor that left oncology for diabetes, Icovamenib (BMF-219) — the covalent menin inhibitor that left oncology for diabetes, Systematic enumeration of menin-inhibitor resistance mechanisms: seven entries written, thirteen candidates rejected with reasons
303. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     ClinicalTrials.gov intervention search for 'icovamenib OR BMF-219', returning all seven registered studies with status, phase, condition and enrolment

     Biomea Fusion Inc. · ClinicalTrials.gov 2026
     used by Biomea Fusion — icovamenib (BMF-219): the covalent menin inhibitor that left oncology for diabetes, Icovamenib (BMF-219) — the covalent menin inhibitor that left oncology for diabetes
304. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.news
     Biomea Fusion pipeline page

     Biomea Fusion, Inc. · biomeafusion.com 2026
     used by Biomea Fusion — icovamenib (BMF-219): the covalent menin inhibitor that left oncology for diabetes
305. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     A Phase 1 Dose Escalation and Expanded Cohort Study of EPZ-5676 in the Treatment of Children With Relapsed/Refractory Leukemias Bearing a Rearrangement of the MLL Gene

     · ClinicalTrials.gov 2026
     used by Bite-back: a DOT1L inhibitor in CALM-AF10 deepens a genome-wide H3K79 deficit that already causes chromosome breaks, No menin-inhibitor or DOT1L-inhibitor trial has ever named this fusion - and the DOT1L trials excluded it by construction, Phase-2 dossier: DOT1L - perfect chemistry, a completed paediatric trial, and a target that has already had its answer, Pinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
306. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
     Provenance scan of 297 mechanism-bearing entries in this graph (folders mechanism-core, histone-marks, interactome1-6, subtraction, subtraction-kmt2a, localization, cells), cycle 49

     Oncofusion knowledge graph, partner-provenance beat · this project 2026
     used by Bite-back: a target nominated from an imported mechanism inherits a fusion the index disease does not have, How much of this graph's mechanism is somebody else's fusion? A counted answer: 54% of fusion-bearing evidence, and 0% from a dedicated human KMT2A::AFF1 experiment, Import class 1: murine MLL-AF9 myeloid transformation, the field's default assay and this graph's largest single import, Import class 2: AHD-dependent mechanism, which cannot transfer because KMT2A::AFF1 has no AHD, Import class 3: the DOT1L-AF10 coiled-coil axis, an MLLT10 mechanism with no counterpart in the index fusion, Import class 4: human myeloid cell lines standing in for infant B-lymphoid disease, The imports that were never necessary: H3K79 methylation in KMT2A::AFF1 has its own literature, and this graph barely cites it, What could refute the menin-interface bet, restated after the provenance audit: three tests, none of which is about menin binding
307. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P55197 (AF10_HUMAN): region 703-784 'Transactivation domain; required for DOT1L-binding' and region 750-778 'Leucine-zipper'

     UniProt Consortium · UniProtKB 2026
     used by Bite-back: a target nominated from an imported mechanism inherits a fusion the index disease does not have, Import class 3: the DOT1L-AF10 coiled-coil axis, an MLLT10 mechanism with no counterpart in the index fusion, KMT2A::AFF1 has no mapped interface to DOT1L: it lacks both structurally defined recruitment modules, and the one complex-level route claimed for it is contested by reconstitution, MLLT10 (AF10) — DOT1L cofactor and KMT2A fusion partner, MLLT10 / AF10 — domain architecture and residue coordinates, Partner import: when a mechanism measured in one fusion is asserted about another
308. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42496646
     Immunoglobulin Depletion and Recovery Following Blinatumomab in Infants With KMT2A-Rearranged ALL

     Vieira Martins M, de Lorenzo P, Kotecha RS, et al. · Pediatric Blood & Cancer 2026
     used by Blinatumomab in infant KMT2A-rearranged ALL — the largest single outcome gain in thirty years, Nine survivors in ten carry a long-term complication, and it is worse the younger they were
309. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.trial_registry
     A Phase II Randomized Study of Blinatumomab With or Without Revumenib for Patients With KMT2A-Translocation B-Cell ALL/ALAL With Persistent Measurable Residual Disease (SWOG); and A Phase IB Trial of Subcutaneous Blinatumomab in Combination With Revumenib for Patients With KMT2A-rearranged ALL (MD Anderson)

     SWOG Cancer Research Network; M.D. Anderson Cancer Center · ClinicalTrials.gov 2026
     used by Blinatumomab in infant KMT2A-rearranged ALL — the largest single outcome gain in thirty years
310. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q8WV28 - B-cell linker protein (Homo sapiens), reviewed

     UniProt Consortium · UniProtKB/Swiss-Prot 2026
     used by BLNK (SLP-65) - the B-cell receptor adapter, in the Q152 AFF1-only core and a known lineage-infidelity marker in myeloid leukaemia
311. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42533114
     PROTAC-mediated targeting of IKKβ and NR4A1 for AML therapy.

     Maharjan CK, Liu Y, ... Zhang W · Oncogene 2026
     used by Blocking IKK/NF-kB in an infant is removing the switch every immune cell runs on — and the class has already failed clinically on exactly that, IKBKB (IKK-beta) — the kinase whose knockdown tracks MLL-leukaemia growth most tightly, and whose inhibitors have failed clinically for a reason, The IKK complex — the unit the eviction experiment actually resolves, because the chemistry that produced the result cannot see the subunits
312. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entries P40424, P40425 and P40426 (PBX1_HUMAN, PBX2_HUMAN, PBX3_HUMAN)

     · UniProtKB/Swiss-Prot 2026
     used by Blot all three PBX proteins in KMT2A::AFF1 B-lineage cells, then delete them singly and in combination - the family has never been tested together in any human leukaemia line
313. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q9NSY1 (BMP2K_HUMAN) feature table and comments, retrieved live 2026-09-24 by q173 via an organism-restricted reviewed gene_exact search that returned exactly one record

     · UniProt Consortium 2026
     used by BMP2K (BIKe) - a nuclear serine/threonine kinase, and the ZNF384 fusion partner that is not a chromatin protein, BMP2K::ZNF384 - a PROTEIN KINASE as the 5' partner, with its catalytic domain retained: the counterexample to 'the partner is always a chromatin coactivator'
314. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Cellosaurus entry CVCL_0064 (MV4-11)

     Bairoch A; SIB Swiss Institute of Bioinformatics · Cellosaurus 2026
     used by BOD1L (BOD1L1) — the top SETD1A co-dependency, and the only candidate in this cycle tested by CRISPR tiling in a KMT2A::AFF1 line, Cell type or the fusion? MV4-11 LOOSENS the confound the graph's own lineage split was built on - but its founding paper calls it biphenotypic, FLT3 is the second largest KMT2A-selective dependency and it is HEAVIER in B-lineage than in myeloid cells, In human KMT2A::AFF1 MV4;11 (monocytic, not B-lineage), the vehicle-versus-vehicle noise floor is 9.4% at the top 0.1% of menin/MLL1 sites and 0.1-2.6% in every other stratum - so cycle 102's borrowed 4-10% band is about right AT THE PEAKS and four to five times too generous everywhere else, and it was applied everywhere, MV4-11 (CVCL_0064) — KMT2A::AFF1, yes, but a ten-year-old's acute monocytic leukaemia with FLT3-ITD, Put the same fusion in two cell types and degrade it, rather than cutting the gene
315. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q8NFC6 (BD1L1_HUMAN), biorientation of chromosomes in cell division protein 1-like 1

     UniProt Consortium · UniProt 2026
     used by BOD1L (BOD1L1) — the top SETD1A co-dependency, and the only candidate in this cycle tested by CRISPR tiling in a KMT2A::AFF1 line
316. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     gnomAD v4 loss-of-function constraint for BOD1L1 (ENSG00000038219) via the gnomAD GraphQL API

     Genome Aggregation Database, Broad Institute · gnomAD 2026
     used by BOD1L (BOD1L1) — the top SETD1A co-dependency, and the only candidate in this cycle tested by CRISPR tiling in a KMT2A::AFF1 line
317. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB data API entry records for 4GQ6 and 2MSR

     · RCSB PDB 2026
     used by Boltz co-folding plan: six complexes worth folding, three that must not be, and the positive controls that decide whether to believe any of it, Phase-2 dossier: the LEDGF/PSIP1 IBD - the best therapeutic index in the interactome, a solved structure, two hotspot residues, and three compounds in the world
318. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     AlphaFold Protein Structure Database prediction API, Q03164 and P51825

     · EMBL-EBI 2026
     used by Boltz co-folding plan: six complexes worth folding, three that must not be, and the positive controls that decide whether to believe any of it, Phase-2 dossier: the fusion junction - a negative feasibility finding, with the one surviving modality named
319. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Cell Ontology term lookup via the EMBL-EBI Ontology Lookup Service OLS4 API

     EMBL-EBI Ontology Lookup Service; Cell Ontology consortium · Database 2026
     used by Bone marrow haematopoietic stem cell — the postnatal comparator, and why the same fusion behaves differently here
320. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42453183
     EZHIP and H3K27me3 are useful diagnostic markers for germinoma and for the germinoma component of germ cell tumors of central nervous system.

     Han L, Hu Z, Zhao L, Teng L, Wang L · Neurooncol Pract 2026
     used by Both MBTD1 fusions end at H3K27me3-poor chromatin, one by excluding PRC2 and one by poisoning it - so the shared endpoint is the phenotype and MBTD1 is the coincidence, EZHIP (CXorf67, CATACOMB) - a protein that jams PRC2 by pretending to be an H3K27M oncohistone, and the 3' payload of MBTD1's other fusion, MBTD1::EZHIP (MBTD1-CXorf67), t(X;17)(p11.2;q21.33) - MBTD1's other fusion, in which a NuA4 subunit delivers an H3K27M-mimicking PRC2 poison to low-grade endometrial stromal sarcoma, One stain, in the right order: H3K27me3 across a low-grade endometrial stromal sarcoma series first, and the three MBTD1::EZHIP blocks second
321. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch db=pubmed via scripts/pubmed_safe.py, 2026-09-23. querytranslations read back from each response: "MBTD1"[Title/Abstract] AND "H3K27me3"[Title/Abstract]; "ZMYND11"[Title/Abstract] AND "H3K27me3"[Title/Abstract]; "EZHIP"[Title/Abstract] AND "H3K27me3"[Title/Abstract]. Counts and id lists persisted to analyses/q123-the-architecture-from-other-parts-analogy/bounded-searches.json

     · NCBI PubMed 2026
     used by Both MBTD1 fusions end at H3K27me3-poor chromatin, one by excluding PRC2 and one by poisoning it - so the shared endpoint is the phenotype and MBTD1 is the coincidence, One stain, in the right order: H3K27me3 across a low-grade endometrial stromal sarcoma series first, and the three MBTD1::EZHIP blocks second
322. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/does-this-deposit-agree-with-itself/run.py and results.json, this repository, computed 2026-09-23 against the cached GEO sample dumps for 32 series

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Bounded absence: every chromatin measurement under a menin inhibitor is in an adult-derived line, and none is in an infant KMT2A::AFF1 model, Re-auditing every cached GEO deposit on three axes after the parser fix: 1,597 comparable title-versus-characteristics pairs across 40 series and 2,620 samples, zero disagreements - and nine series including the one this cycle's beats rest on are abstentions rather than passes, because a 0-of-0 is not a clean bill of health, RETRACTED: 'Sixteen percent of GEO sample labels contradict their own metadata' was an off-by-one in my own parser - GEO's brief format puts the title before the accession, so every sample was scored against the next sample's title; the true rate is ZERO, the deposit named as worst offender has none, and the same bug made me 'correct' an agent whose accessions were right
323. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed esearch counts for four stated queries on infant leukaemia dependency screening, run 2026-09-19

     Oncofusion knowledge graph, Q21 index-fusion-models beat · NCBI Entrez eutils (esearch, db=pubmed) 2026
     used by Bounded absence: no published CRISPR or dependency screen in any infant-derived KMT2A::AFF1 model
324. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, queries and counts run 2026-09-19

     NCBI · NCBI E-utilities 2026
     used by Bounded absence: not one menin-inhibitor study, trial or preclinical paper has ever recorded reciprocal fusion status, Enumeration of the reciprocal KMT2A fusion literature: what was searched, found, rejected and could not be read
325. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST core record and NCBI elink for PMID 34551297

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure with a near-miss worth recording: a GEO series matching the topic exactly belongs to a different paper
326. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 41685902
     Development of DOT1L-Targeted Protein Degraders for Treating MLL-r Leukemia.

     Quan S, Unno K, Gupta DG, Yang W, Guo Q, Dufficy MF, Kerber MM, Elmashae AA, Pan H, Zhu JL, Schiltz GE, Abdulkadir SA · J Med Chem 2026
     used by Bounded failure: 'menin inhibitor-resistant cells' is a drug entry's central claim and neither the paper nor any repository defines it, Degrade DOT1L, inhibit DOT1L, or leave a dead enzyme in place - run all three against the same cells and the same marrow, DOT1L does things that have nothing to do with methylation - which is why blocking its active site was never going to remove all of it, DOT1L was not refuted - the wrong thing was blocked, at a dose nobody finished escalating, DOT1L808 - a 5 nM DOT1L degrader that cleared an orthotopic leukaemia and still works after menin inhibitors fail, MS2133 - the first molecule to degrade DOT1L rather than inhibit it
327. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST core record, NCBI elink and GEO keyword search for PMID 41685902

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure: 'menin inhibitor-resistant cells' is a drug entry's central claim and neither the paper nor any repository defines it
328. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC fullTextXML and NCBI efetch db=pmc responses for PMC7566887

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure: a PMCID that exists and a body that does not - the MLL amino-terminus transformation paper is unreadable by both PMC routes
329. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI elink pubmed-to-gds for PMID 17671196 and esearch db=gds author query

     · NCBI Entrez Programming Utilities 2026
     used by Bounded failure: a PMCID that exists and a body that does not - the MLL amino-terminus transformation paper is unreadable by both PMC routes
330. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC fullTextXML and NCBI efetch db=pmc responses for PMC6234363

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure: the CDK4/6 cell lines behind two therapeutic entries are unrecoverable although the article is marked free access
331. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI elink pubmed-to-gds and esearch db=gds keyword query for PMID 30381403

     · NCBI Entrez Programming Utilities 2026
     used by Bounded failure: the CDK4/6 cell lines behind two therapeutic entries are unrecoverable although the article is marked free access
332. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST core record and NCBI elink pubmed-to-gds for PMID 16239140

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure: the founding menin-cofactor paper names no fusion, has no PMC record and deposited nothing - and three entries depend on it
333. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 42537480
     Influence of IGF2BP3 on t(4;11) fusion mRNA and protein abundance.

     Hanewald T, Benz T, Marschalek R · Leuk Res 2026
     used by Bounded failure: the one unresolved source that is actually about the index fusion, Hypothesis: menin inhibition does not merely displace KMT2A::AFF1 from chromatin, it destabilises it — the evicted fusion is degraded, not parked, IGF2BP3 — an RNA-binding protein required for MLL-Af4 leukaemogenesis in mice that CRISPR barely touches in human lines, IGF2BP3 works in cytoplasmic RNA granules - the only compartment in this beat where a drug is already pointed, and nobody has looked at it in a leukaemia cell, Infants whose leukaemia expresses the reciprocal fusion survive far better - 62.4% versus 11.7% event-free at four years, LIN28B and IGF2BP3 are co-regulated members of ONE onco-fetal module in infant leukaemia - which strengthens the graph's existing scepticism about IGF2BP3 rather than weakening it, Put both fusion proteins in the same cell and you get something neither produces alone, RNA-binding proteins set how much KMT2A::AFF1 protein a cell makes from its own fusion transcript, SIAH1/SIAH2: the disposal route that destroys wild-type AF4 and that the reciprocal fusion escapes, Spliced KMT2A fusions: breaks upstream of the partner's exon 1, and the reciprocal-free natural experiment, The reciprocal fusion (AFF1::KMT2A and friends): a second oncoprotein that opens chromatin for the first
334. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST core record and NCBI elink for PMID 42537480

     · Europe PMC and NCBI Entrez Programming Utilities 2026
     used by Bounded failure: the one unresolved source that is actually about the index fusion
335. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI esearch db=gds keyword query 'IGF2BP3 AND leukemia'

     · NCBI Gene Expression Omnibus 2026
     used by Bounded failure: the one unresolved source that is actually about the index fusion
336. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Europe PMC REST core record for EXT_ID:21030982 - pmcid null, isOpenAccess N, inPMC N, hasPDF N; fullTextUrlList gives only a subscription-required DOI link

     · Europe PMC 2026
     used by Bounded unreadable: the only affinity purification of a SET-retaining KMT2A-derived fusion complex fails all four routes, and the molecule is the reciprocal product
337. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI PMC ID converter for PMID 21030982 - status error, 'Identifier not found in PMC'; and elink dbfrom=pubmed&db=pmc returns only pubmed_pmc_refs (48 citing articles), no pubmed_pmc link

     · NCBI E-utilities 2026
     used by Bounded unreadable: the only affinity purification of a SET-retaining KMT2A-derived fusion complex fails all four routes, and the molecule is the reciprocal product
338. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Publisher HTML page for Leukemia 25:135-144 (doi 10.1038/leu.2010.249) - HTTP 200, 368,813 bytes, 18,543 characters of stripped text consisting of abstract, paywall notice, figure list and reference list, with no Results or Methods

     · Springer Nature 2026
     used by Bounded unreadable: the only affinity purification of a SET-retaining KMT2A-derived fusion complex fails all four routes, and the molecule is the reciprocal product
339. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ENCODE Portal SUV39H1 target and experiment queries; ChIP-Atlas hg38 target table SUV39H1.5.tsv (26 genes, columns SRX7030945-48, all MCF-7)

     ENCODE Project Consortium; Oki S, Ohta T et al. · ENCODE Portal / ChIP-Atlas (DBCLS) 2026
     used by Bounded with every archive and every route named: after both public SUV39H1 deposits were read this cycle, there is still NO SUV39H1 occupancy measurement in any haematopoietic cell, any KMT2A-rearranged cell, or any damage state - in ENCODE (three routes, zero), ChIP-Atlas (one deposit, MCF-7 only) or GEO (two deposits, glioma and breast), The one experiment that would close the SUV39H1 arm of this graph's H3K9 argument: CST #8729S CUT&RUN in a CD34+ progenitor and in SEM or RS4;11, with an IgG arm and a SUV39H1-degron arm that neither existing deposit has - because after this cycle the enzyme has been measured in exactly two human cell types and neither is a blood cell, Three archives, one answer: SUV39H1 has no ENCODE target record under any label (three routes, all empty), ChIP-Atlas's entire hg38 SUV39H1 holding IS the four GSE139196 MCF-7 samples re-indexed, and KMT2A is not among the 26 genes ChIP-Atlas calls SUV39H1 targets
340. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     EGA public metadata API https://metadata.ega-archive.org/studies paged by limit=1000 and offset, 10,649 records, retrieved 2026-09-22; CEEHRC/CEMT reference-epigenome study EGAS00001000552 and IHEC DEEP release EGAS00001001937 read for the core-mark panel

     · European Genome-phenome Archive, EMBL-EBI / CRG 2026
     used by Bounded, with every archive named: H3K9me1 and H3K9me2 in a CD34+ progenitor are absent from BLUEPRINT, EGA, Roadmap/Human Epigenome Atlas, the IHEC Data Portal and ENCODE - 49,663 IHEC datasets carry 178 assay labels and H3K9me2 is not one of them - and the absence is a consortium artefact, because GEO has had the CD34+ H3K9me2 data since 2017
341. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ENCODE portal search API, type=Experiment&assay_title=Histone+ChIP-seq&target.label=H3K9me1|H3K9me2|H3K9me3 with biosample_ontology and donor organism fields, retrieved 2026-09-22

     · ENCODE Project, encodeproject.org 2026
     used by Bounded, with every archive named: H3K9me1 and H3K9me2 in a CD34+ progenitor are absent from BLUEPRINT, EGA, Roadmap/Human Epigenome Atlas, the IHEC Data Portal and ENCODE - 49,663 IHEC datasets carry 178 assay labels and H3K9me2 is not one of them - and the absence is a consortium artefact, because GEO has had the CD34+ H3K9me2 data since 2017
342. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P38398 (BRCA1_HUMAN) - Breast cancer type 1 susceptibility protein, Homo sapiens

     UniProt Consortium · UniProt Knowledgebase 2026
     used by BRCA1 (P38398, 1,863 aa) - named in 19 entries of this graph and never written: the HR mediator that CTCF and the H3K9 machinery are now reported to license at a break, and the only cascade member with a human therapy-related-leukaemia signal after topoisomerase II therapy
343. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ENCODE Portal experiment search, type=Experiment, assay_title=TF ChIP-seq, status=released, per target label

     ENCODE Project Consortium · ENCODE Portal 2026
     used by BRCA1 (P38398, 1,863 aa) - named in 19 entries of this graph and never written: the HR mediator that CTCF and the H3K9 machinery are now reported to license at a break, and the only cascade member with a human therapy-related-leukaemia signal after topoisomerase II therapy, No BRCA1, BARD1, EXO1, DNA2, HP1-gamma or KDM5A has ever been localised at the KMT2A locus, in any cell type: 21 queries, every querytranslation read back, 0 on-topic records - and the absence splits THREE ways by class, which is what decides what it costs to close
344. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     ENCODE BRCA1 ChIP-seq fold-change-over-control bigWigs ENCFF690UED (ENCSR000EDB, HeLa-S3), ENCFF620MRE (ENCSR000EBX, H1), ENCFF133FPU (ENCSR343RJR, HepG2), ENCFF191NNV (ENCSR000EDY, HepG2), ENCFF493XJG (ENCSR223MLH, K562), ENCFF017KXT (ENCSR000DZS, GM12878), read 2026-09-22 against CTCF anchor sets ENCFF533TKY, ENCFF497ZGV, ENCFF695QSJ, ENCFF041ODC and ENCFF796WRU

     ENCODE Project Consortium · ENCODE Portal 2026
     used by BRCA1 at the KMT2A exon-11 CTCF anchor is real, cell-type-restricted, and survives the summit-matched null Q35 asked for: 93.7th and 92.3rd percentile of same-cell CTCF anchors in HeLa-S3 and H1, rising to 95.2 and 95.7 under width-and-TSS matching - and 34.0 and 45.1 in the only two blood lines, where ENCODE's 300,000-peak pre-IDR lists are also empty, Every peak file of all 13 ENCODE BRCA1, CBX3 and CBX1 experiments opened at the exon-11 element, not just the two Q37 reached: BRCA1 is there in BOTH pre-IDR replicates of THREE experiments - HeLa-S3, H1 and HepG2, global IDR 0.059, 0.577 and 0.482 - and it is a clean zero in both K562 experiments and in GM12878 against 300,000-peak pre-IDR lists
345. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR gene-effect (Chronos) for BRCA2, served by the Open Targets Platform GraphQL API

     Broad Institute DepMap / Open Targets Platform · Open Targets Platform 26.06 (meta.dataVersion year 26 month 06), fetched 2026-09-22 2026
     used by BRCA2 (P51587, 3,418 aa) - the RAD51 loader that receives the resected end BRCA1 licensed, named in 10 entries and never written; DepMap mean -0.519, and its 12-record KMT2A intersection is entirely germline-panel and pan-cancer sequencing with no mechanism
346. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed E-utilities esearch: (MLL[tiab] OR KMT2A[tiab]) AND (BRD2[tiab] OR BRD3[tiab]), run 2026-09-20, count 7, all seven title-checked by efetch (37594294, 36170346, 35832114, 33956058, 30952871, 29240787, 21964340)

     NCBI E-utilities · PubMed 2026
     used by BRD2 - a bounded absence: zero KMT2A papers name it, it is inhibited by every BET compound used in this disease, and it is the paralogue most likely to carry the toxicity
347. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed E-utilities esearch: (MLL[tiab] OR KMT2A[tiab]) AND ("BET bromodomain"[tiab] OR "BET protein"[tiab] OR "BET proteins"[tiab]), run 2026-09-20, count 17, all titles checked by efetch

     NCBI E-utilities · PubMed 2026
     used by BRD2 - a bounded absence: zero KMT2A papers name it, it is inhibited by every BET compound used in this disease, and it is the paralogue most likely to carry the toxicity
348. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P25440 (BRD2_HUMAN), Bromodomain-containing protein 2

     UniProt Consortium · UniProtKB, retrieved 2026-09-20 2026
     used by BRD2 - a bounded absence: zero KMT2A papers name it, it is inhibited by every BET compound used in this disease, and it is the paralogue most likely to carry the toxicity
349. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q15059 (BRD3_HUMAN), Bromodomain-containing protein 3

     UniProt Consortium · UniProtKB, retrieved 2026-09-20 2026
     used by BRD3 - named in the founding BET paper's own mechanism and in one title since, and not once separated from BRD4 by any experiment in this disease
350. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q15059 (BRD3_HUMAN), SUBCELLULAR LOCATION comment with evidence codes and sources plus the feature table, fetched live 2026-09-22

     UniProt Consortium · Databases 2026
     used by BRD3 - named in the founding BET paper's own mechanism and in one title since, and not once separated from BRD4 by any experiment in this disease
351. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O60885 PDB cross-references; RCSB PDB entries 3MXF, 2OSS and 4LYS verified against the RCSB data API

     UniProt Consortium; RCSB Protein Data Bank · Databases 2026
     used by BRD4 — the best-drugged protein in this beat, and the one whose clinical result is already in and disappointing
352. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9H8M2 (BRD9_HUMAN), Bromodomain-containing protein 9

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by BRD9 — dispensable in the average cell, rank 39 of 17,728 in KMT2A-rearranged lines, and with an on-target proof most epigenetic targets never get, BRD9::NUTM1, t(5;15)(p15.33;q14) - one fusion is enough to make infant B-ALL in a mouse, and it is the one that behaves best in a child
353. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch record counts for BRD9 against KMT2A, MLL and AML terms, run for this beat

     · NCBI PubMed, queried 2026-09-19 2026
     used by BRD9 — dispensable in the average cell, rank 39 of 17,728 in KMT2A-rearranged lines, and with an on-target proof most epigenetic targets never get
354. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Frequently Asked Questions: Breakthrough Therapies

     US Food and Drug Administration · FDA (FDASIA implementation) 2026
     used by Breakthrough therapy and accelerated approval: the menin field took the fast review and declined the shortcut
355. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Accelerated Approval Program

     US Food and Drug Administration · FDA 2026
     used by Breakthrough therapy and accelerated approval: the menin field took the fast review and declined the shortcut
356. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paperPMID 42215936
     Extreme hyperleukocytosis and blueberry muffin skin lesions as the initial presentation of neonatal leukemia: a case report.

     Bustamante-Ordonez MA, Macavilca-Toribio MA, Arenas-Vasquez JR, Roman-Lazarte V, Zavaleta-Corvera C, Samanez-Obeso A · BMC Pediatr 2026
     used by Bringing down a white count of half a million in a four-kilogram baby, Leukaemia diagnosed in the first month of life: a different presentation and a different prognosis, Pallor and anaemia in infants: the number this graph quotes does not exist for babies, Rash: the one presenting feature that gets commoner the younger the baby is
357. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/directed-interactome-coverage-recheck/run.py and results.json, this repository, re-run 2026-09-24, with the three uncovered accessions resolved against the UniProt REST API

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by BRINP1 (DBC1/DBCCR1/FAM5A) - the last bait-supported partner of the six core proteins without an entry, a neural-specific BMP/retinoic-acid-inducible protein pulled down with DOT1L as bait in one quantitative AP-MS experiment and nowhere else
358. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q9BX63

     · UniProtKB/Swiss-Prot 2026
     used by BRIP1 — downstream effector; rank 22 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all
359. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed counted absence search for BRIP1 against KMT2A/MLL, run 2026-09-20 with aliases and title-checked

     · PubMed (NCBI) 2026
     used by BRIP1 — downstream effector; rank 22 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all
360. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9BRD0 (BUD13_HUMAN) features, comments, PDB cross-references and subcellular-location evidence codes; HGNC:28199 via rest.genenames.org; Ensembl ENSG00000137656 and ENSG00000118058 GRCh38 coordinates via rest.ensembl.org; retrieved 2026-09-22

     · UniProt Consortium / HGNC / Ensembl 2026
     used by BUD13 - a RES-complex splicing factor 1.66 Mb from KMT2A on 11q23.3, whose one chromatin paper puts it in the soluble fraction
361. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     IntAct REST findInteractions/Q9BRD0 (673 indexed interactions), filtered to Q03164/O00255/O75475/Q9UHB7/Q8TEK3/P51825/Q03111/P42568/P55199/P55197 with ac, binaryInteractionId, experimentalRole, detectionMethod and hostOrganism read per row; retrieved 2026-09-22

     · IntAct molecular interaction database (EMBL-EBI) 2026
     used by BUD13 - a RES-complex splicing factor 1.66 Mb from KMT2A on 11q23.3, whose one chromatin paper puts it in the soluble fraction
362. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     analyses/depmap-kmt2a-selective-scan-full/results.json - 17,728 protein-coding genes, 11 Cellosaurus-annotated KMT2A-rearranged versus 43 other leukaemia lines, computed 2026-09-19

     · DepMap (Broad Institute), via this repository's scan 2026
     used by BUD13 - a RES-complex splicing factor 1.66 Mb from KMT2A on 11q23.3, whose one chromatin paper puts it in the soluble fraction
363. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch, db=pubmed, querytranslation read back from every response, run 2026-09-22

     · NCBI PubMed 2026
     used by BUD13 - a RES-complex splicing factor 1.66 Mb from KMT2A on 11q23.3, whose one chromatin paper puts it in the soluble fraction
364. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/has-this-corpus-already-measured-this/run.py and results.json, this repository, computed 2026-09-24 against the committed node files

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Building the instrument that asks whether this corpus has already measured something: it fails, 85% of comparable pairs flag as disagreements and the one real conflict never surfaces - because measurements live in prose, so the fix is a typed reporting field and it is now in the standing brief rather than in an open question
365. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41779961
     Refining busulfan exposure enhances pediatric ALL HSCT outcomes: insights from the International FORUM study.

     Ben Hassine K, Gloor Y, Dupanloup I, Uppugunduri CRS, Mlakar V, Gonzales F, Guengoer T, Ifversen M, et al. · Blood Advances 2026
     used by Busulfan - the alkylating agent that replaces radiation in infant conditioning, and whose labelled paediatric target range is below the range the modern trial data support, Mixed donor chimerism under three years of age - the same drug exposure ablates less marrow in a younger child, and dosing cannot fix it, Sinusoidal obstruction syndrome after infant conditioning - an age signal, a regulatory exposure threshold, and a prophylaxis that failed its randomised trial, Total body irradiation and the age-four boundary - the best conditioning in childhood ALL has never been tested in the children who most need a transplant, Weight or surface area - the Interfant-06 conditioning week uses both units in eight days, and the one measured correction in the literature went the wrong way
366. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q4AC94 (C2CD3_HUMAN) - C2 domain-containing protein 3 (canonical isoform 5, 2,353 aa)

     UniProt Consortium · UniProtKB 2026
     used by C2CD3 - the centriole-elongation regulator fused to KMT2A exon 8 in an MPNST, predicted OUT OF FRAME, and the partner whose RefSeq transcript is 390 residues shorter than UniProt's canonical
367. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RefSeq NM_015531.6 - Homo sapiens C2 calcium dependent domain containing 3 (C2CD3), mRNA - NP_056346.3, 1,963 aa

     NCBI RefSeq · NCBI Nucleotide 2026
     used by C2CD3 - the centriole-elongation regulator fused to KMT2A exon 8 in an MPNST, predicted OUT OF FRAME, and the partner whose RefSeq transcript is 390 residues shorter than UniProt's canonical
368. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset
     DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan)

     Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
     used by C2CD3 - the centriole-elongation regulator fused to KMT2A exon 8 in an MPNST, predicted OUT OF FRAME, and the partner whose RefSeq transcript is 390 residues shorter than UniProt's canonical
369. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record for C4orf36 (HGNC:28386) and UniProtKB search gene_exact:C4orf36 AND organism_id:9606

     HUGO Gene Nomenclature Committee; UniProt Consortium · HGNC / UniProt 2026
     used by C4orf36 - a protein-coding gene with three current TSSs spanning 43,782 bp, two of them inside or beside AFF1's promoter window, and the project's table has room for one
370. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UCSC Genome Browser hg38 ncbiRefSeqCurated track, chr4:86,800,000-87,000,000

     UCSC Genome Browser, api.genome.ucsc.edu · UCSC Genome Browser 2026
     used by C4orf36 - a protein-coding gene with three current TSSs spanning 43,782 bp, two of them inside or beside AFF1's promoter window, and the project's table has room for one, REFUTED THE SAME CYCLE: window displacement against the current annotation is 1.83%, not systematic - and the real blind spot it hid is one row per symbol at 14.69%, RUN, by Q156, within the hour: the whole-table displacement audit - drift is 1.83%, the real blind spot is one row per symbol at 14.69%, and the case that motivated it is not displaced
371. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Ensembl REST lookup/id for ENSG00000163633 and ENSG00000285458, and overlap/region/human/4:86915000-86945000, GRCh38

     Ensembl REST API, rest.ensembl.org · Ensembl 2026
     used by C4orf36 - a protein-coding gene with three current TSSs spanning 43,782 bp, two of them inside or beside AFF1's promoter window, and the project's table has room for one
372. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9Y6J0 (CABIN_HUMAN), Calcineurin-binding protein cabin-1, 2220 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CABIN1 - a 2,220-residue calcineurin inhibitor that is also a full member of the H3.3 chaperone complex, and nobody has asked which job matters in blood
373. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.reviewPMID 41532548
     PICALM::MLLT10 translocated leukemia.

     Cullen JM, et al. · FEBS Lett 2026
     used by CALM-AF10 needs BMI1 - and so does MLL-AF9, so this is a shared polycomb dependency and not a point of difference (title corrected 2026-09-21), CALM-AF10 was meant to separate the menin axis from the HOXA programme - and the indirect evidence has now turned against separability, because KMT2A-free HOXA lesions keep proving menin-DEPENDENT, How common PICALM::MLLT10 is, and in which disease - 65% T-ALL, 27% AML, and invisible to karyotype half the time, Nobody has ever published a menin experiment in CALM-AF10 - counted, controlled and dated, Outcome in PICALM::MLLT10: children with ALL do reasonably well, children with AML do badly, and transplant does not rescue them, PICALM::MLLT10 (CALM-AF10), t(10;11)(p13;q14) - the DOT1L-dependent HOXA fusion that contains no KMT2A
374. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 6CKN, 6CKO and 6JN2 - DOT1L coiled-coil in complex with the MLLT10/AF10 OM-LZ

     · RCSB PDB 2026
     used by CALM-AF10 reaches DOT1L through exactly the same OM-LZ interface as KMT2A::MLLT10 - the recruitment transfers, the addressing does not, MLLT10 / AF10 — domain architecture and residue coordinates, Where the break falls, and the three incompatible numbering frames that make it easy to get wrong
375. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed E-utilities esearch and Europe PMC REST counts for CALM-AF10 / PICALM / MLLT10 / AF10 against a menin and menin-inhibitor name list

     · NCBI PubMed / Europe PMC 2026
     used by CALM-AF10 was meant to separate the menin axis from the HOXA programme - and the indirect evidence has now turned against separability, because KMT2A-free HOXA lesions keep proving menin-DEPENDENT, Nobody has ever published a menin experiment in CALM-AF10 - counted, controlled and dated, PICALM::MLLT10 (CALM-AF10), t(10;11)(p13;q14) - the DOT1L-dependent HOXA fusion that contains no KMT2A, Run the menin inhibitor against CALM-AF10 next to a KMT2A-rearranged control - the missing two-week experiment, Why this one rare leukaemia was thought to decide an argument about a drug used in a commoner one - and why it decides less than the title says (corrected 2026-09-21)
376. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 42246204
     BCL-2 and MCL1 co-inhibition enhances glucocorticoid therapy in preclinical models of infant KMT2A-rearranged acute lymphoblastic leukaemia.

     Karsa M, Karsa A, Spurling D, ..., Somers K · Br J Haematol 2026
     used by Cardiac troponin release on MCL-1 inhibition - reported for two structurally distinct compounds and described by one developer as a likely CLASS effect, which makes it an essential-function ceiling rather than a chemistry problem, Kotecha group, Perth: the infant KMT2A-rearranged laboratory this graph had cited once, MCL1 - named in 34 entries of this graph with no entry of its own until now; a mitochondrial survival protein high in infant KMT2A-rearranged ALL, druggable, and so far undruggable in patients, MCL1 plus BCL-2 plus dexamethasone: a triple combination that matched induction chemotherapy in infant PDX models, Venetoclax in KMT2A-rearranged leukemia — the apoptotic partner every other drug needs
377. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry O75604 (UBP2_HUMAN), Ubiquitin carboxyl-terminal hydrolase 2, active sites and domain boundaries

     UniProt Consortium; InterPro (EMBL-EBI) · UniProtKB / InterPro 2026
     used by Catalytic-dead UCH test: does KMT2A::USP2 need its deubiquitinase activity to transform?, KMT2A::USP2, inv(11)(q23.3q23.3) - the minor-BCR fusion that keeps the PHD fingers, USP2 (ubiquitin carboxyl-terminal hydrolase 2) - the ninth commonest KMT2A partner and the only one that is a deubiquitinase
378. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     UniProtKB O75081 (CBFA2T3_HUMAN): sequence length, domain, zinc-finger, DNA-binding, repeat, motif and region features, and PDB cross-references

     UniProt Consortium · UniProtKB 2026
     used by CBFA2T3 (ETO2 / MTG16) domain architecture
379. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     RCSB PDB: search by UniProt accession Q9BZE0 (GLIS2) returns zero entries; O75081 (CBFA2T3) returns 9DE2 (ETO2 MYND bound to MPPL peptide from GATAD2A); paralogue entries 1WQ6 (NHR2 tetramer of AML1-ETO) and 2ODD (AML1-ETO MYND with SMRT)

     RCSB Protein Data Bank · RCSB PDB 2026
     used by CBFA2T3::GLIS2, inv(16)(p13.3q24.3)
380. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q13951 (PEBB_HUMAN): reviewed sequence, length and positional feature annotations

     UniProt Consortium · UniProtKB/Swiss-Prot 2026
     used by CBFB — domain architecture and residue coordinates
381. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro/Pfam/CATH-Gene3D/SUPERFAMILY domain matches for UniProt Q13951

     InterPro Consortium; EMBL-EBI · InterPro 2026
     used by CBFB — domain architecture and residue coordinates
382. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     PDB 1CL3 'MOLECULAR INSIGHTS INTO PEBP2/CBF-SMMHC ASSOCIATED ACUTE LEUKEMIA REVEALED FROM THE THREE-DIMENSIONAL STRUCTURE OF PEBP2/CBF BETA', with 1E50 'AML1/CBFbeta complex' and 1H9D 'Aml1/cbf-beta/dna complex' (all three titles retrieved and verified at RCSB)

     RCSB Protein Data Bank (entry metadata retrieved and title verified) · RCSB PDB 2026
     used by CBFB — domain architecture and residue coordinates
383. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     PDB 8CX0 'Cryo-EM structure of human APOBEC3G/HIV-1 Vif/CBFbeta/ELOB/ELOC monomeric complex' and 6P59 'Crystal structure of SIVrcm Vif-CBFbeta-ELOB-ELOC complex': observed-residue ranges for the CBFB chains

     RCSB Protein Data Bank (entry titles and instance observed-residue records retrieved and read back) · RCSB PDB 2026
     used by CBFB — domain architecture and residue coordinates
384. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     RCSB PDB: search by UniProt accession P35749 (MYH11) returns one entry, 9FU2 (Smooth Muscle Myosin II in complex with MT-228); Q13951 (CBFB) structures appear within the RUNX1 heterodimer entries

     RCSB Protein Data Bank · RCSB PDB 2026
     used by CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)
385. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry P35520 (CBS_HUMAN), Cystathionine beta-synthase

     UniProt Consortium · UniProt 2026
     used by CBS (P35520) - the largest fall in the whole panel (-4.07), a token that collides with Cajal bodies AND CTCF binding sites, and a germline phenotype whose hallmark is thrombosis
386. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI E-utilities esearch against PubMed, 2026-09-22, fusion operand set (KMT2A[tiab] OR MLL[tiab] OR "MLL-AF4"[tiab] OR "KMT2A::AFF1"[tiab] OR "MLL-rearranged"[tiab] OR "t(4;11)"[tiab] OR "MLL-AF9"[tiab] OR "MLL-ENL"[tiab]), each term separately tagged and every cytogenetic term quoted

     · PubMed 2026
     used by CBS (P35520) - the largest fall in the whole panel (-4.07), a token that collides with Cajal bodies AND CTCF binding sites, and a germline phenotype whose hallmark is thrombosis
387. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P83916 (CBX1), reviewed human entry, fetched 2026-09-22 by accession

     UniProt Consortium · UniProtKB 2026
     used by CBX1 / HP1-beta (P83916, 185 aa) - the HP1 paralogue with the fitness cost and no role in this graph's cascade: DepMap mean -0.349 with 16.6% of 1,257 lines below -0.5, four times deeper than HP1-gamma or HP1-alpha, and zero papers place it anywhere near KMT2A, Decompose the CBX1 DepMap screen to single guides before anyone quotes -0.349 as biology: a no-wet-work, one-afternoon check that would either kill the HP1 paralogue asymmetry or make it the only HP1 question in this graph with a phenotype, The HP1 paralogue asymmetry is a guide-cross-reactivity artefact until somebody shows it is not - and if it is real, HP1-beta is doing a job the other two cannot, which is the only HP1 question in this graph with a fitness phenotype attached
388. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR gene-effect (Chronos) for CBX1, served by the Open Targets Platform GraphQL API

     Broad Institute DepMap / Open Targets Platform · Open Targets Platform 26.06 (meta.dataVersion year 26 month 06), fetched 2026-09-22 2026
     used by CBX1 / HP1-beta (P83916, 185 aa) - the HP1 paralogue with the fitness cost and no role in this graph's cascade: DepMap mean -0.349 with 16.6% of 1,257 lines below -0.5, four times deeper than HP1-gamma or HP1-alpha, and zero papers place it anywhere near KMT2A, Decompose the CBX1 DepMap screen to single guides before anyone quotes -0.349 as biology: a no-wet-work, one-afternoon check that would either kill the HP1 paralogue asymmetry or make it the only HP1 question in this graph with a phenotype, The HP1 paralogue asymmetry is a guide-cross-reactivity artefact until somebody shows it is not - and if it is real, HP1-beta is doing a job the other two cannot, which is the only HP1 question in this graph with a fitness phenotype attached
389. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     ENCODE Portal: released human CBX1 TF ChIP-seq experiments ENCSR948QLZ (K562) and ENCSR819WZE (HepG2), signal and peak files, analysed 2026-09-22

     ENCODE Project Consortium · ENCODE Portal 2026
     used by CBX1 / HP1-beta (P83916, 185 aa) - the HP1 paralogue with the fitness cost and no role in this graph's cascade: DepMap mean -0.349 with 16.6% of 1,257 lines below -0.5, four times deeper than HP1-gamma or HP1-alpha, and zero papers place it anywhere near KMT2A
390. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q13185 (CBX3_HUMAN) - Chromobox protein homolog 3 / HP1 homolog gamma, Homo sapiens

     UniProt Consortium · UniProt Knowledgebase 2026
     used by CBX3 / HP1-gamma (Q13185, 183 aa) - the EUCHROMATIC HP1 paralogue, and the one CTCF actually prefers: the reader that carries BRCA1/BARD1 onto H3K9-methylated break chromatin, flat in DepMap, and a fusion partner in one KMT2A::AFDN AML patient
391. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q13185, P83916 and P45973 - the three human HP1 paralogues, fetched in one gene_exact query set

     UniProt Consortium · UniProt Knowledgebase 2026
     used by CBX3 / HP1-gamma (Q13185, 183 aa) - the EUCHROMATIC HP1 paralogue, and the one CTCF actually prefers: the reader that carries BRCA1/BARD1 onto H3K9-methylated break chromatin, flat in DepMap, and a fusion partner in one KMT2A::AFDN AML patient
392. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ENCODE Portal experiment search, type=Experiment, assay_title=TF ChIP-seq, status=released, target.label=CBX3

     ENCODE Project Consortium · ENCODE Portal 2026
     used by CBX3 / HP1-gamma (Q13185, 183 aa) - the EUCHROMATIC HP1 paralogue, and the one CTCF actually prefers: the reader that carries BRCA1/BARD1 onto H3K9-methylated break chromatin, flat in DepMap, and a fusion partner in one KMT2A::AFDN AML patient
393. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P45973 (CBX5_HUMAN) - Chromobox protein homolog 5 / HP1 alpha, Homo sapiens; with HGNC:1555

     UniProt Consortium; HUGO Gene Nomenclature Committee · UniProt Knowledgebase / HGNC 2026
     used by CBX5 / HP1-alpha (P45973, 191 aa) - the H3K9me3 reader that recruits the writer, the writer's destroyer and the second writer; and the one member of that circuit with a published pro-leukaemic role, as part of an MLL leukaemia stem-cell programme that immortalises myeloid progenitors WITHOUT Hoxa or Meis
394. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42055326
     HIRA-SETDB1-H3K9me3 axis regulates chromatin architecture in leukemia cells.

     Shirude MB, Devarajan A, Sahu SA, Mukherjee A, et al. · The Journal of Biological Chemistry 2026
     used by CBX5 / HP1-alpha (P45973, 191 aa) - the H3K9me3 reader that recruits the writer, the writer's destroyer and the second writer; and the one member of that circuit with a published pro-leukaemic role, as part of an MLL leukaemia stem-cell programme that immortalises myeloid progenitors WITHOUT Hoxa or Meis, HIRA - the chaperone that puts new H3.3 into transcribed gene bodies, and therefore the machine that runs this graph's clearance clock
395. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41712379
     Local heterochromatin enrichment promotes telomere clustering and PML nuclear body assembly at telomeres.

     Taylor ER, Proctor B, Vaurs M, Wu G, et al. · Cell Reports 2026
     used by CBX5 / HP1-alpha (P45973, 191 aa) - the H3K9me3 reader that recruits the writer, the writer's destroyer and the second writer; and the one member of that circuit with a published pro-leukaemic role, as part of an MLL leukaemia stem-cell programme that immortalises myeloid progenitors WITHOUT Hoxa or Meis
396. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 5EQ0, 3I91 and 2N4Q, with per-entity UniProt accessions and modelled residue ranges resolved against the RCSB search and data APIs

     RCSB Protein Data Bank · RCSB PDB 2026
     used by CBX8 — the one place in this interactome where two good papers flatly disagree, and DepMap takes a side
397. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, target CHEMBL3232684 (Chromobox protein homolog 8): target_search, get_bioactivity and get_mechanism queries

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by CBX8 — the one place in this interactome where two good papers flatly disagree, and DepMap takes a side
398. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9HC52 (CBX8_HUMAN), SUBCELLULAR LOCATION comment with evidence code and source, and gene synonym list, fetched live 2026-09-22

     UniProt Consortium · Databases 2026
     used by CBX8 — the one place in this interactome where two good papers flatly disagree, and DepMap takes a side
399. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch db=pubmed: CBX3[tiab] AND KMT2A[tiab] = 1 (PMID 39823827); CBX3[tiab] AND MLL[tiab] = 1 (same record); HP1gamma[tiab] AND MLL[tiab] = 0; "HP1-gamma"[tiab] AND MLL[tiab] = 0; "HP1 gamma"[tiab] AND MLL[tiab] = 0

     · NCBI PubMed 2026
     used by CCDC32::CBX3, a previously unreported second fusion found alongside KMT2A::AFDN in a relapsing AML - and the only record in PubMed that pairs CBX3 with KMT2A at all
400. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P30281 (CCND3_HUMAN) - SUBCELLULAR LOCATION comment block with evidence codes

     · UniProt Consortium 2026
     used by CCND3 (cyclin D3) — the scan's rank-4 hit, and the only one of the four with a drug class already in children
401. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 3BLH, 6Z45 and 7NWK, verified against the RCSB data API for title, experimental method and resolution

     RCSB Protein Data Bank · RCSB PDB 2026
     used by CCNT1 / cyclin T1 — the regulatory half of P-TEFb, and the one part of the elongation core with any dependency variance
402. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC REST records fetched 2026-09-20: CDK13 (HGNC symbol, 7p14.1, previous symbol CDC2L5, aliases CHED/CDC2L/KIAA1791), CDK12 (17q12, previous symbol CRKRS, aliases CRK7/CRKR/KIAA0904), CCNT2 (2q21.3, no aliases), CSNK2A2 (16q21, aliases CSNK2A1 and CK2alpha-prime)

     HUGO Gene Nomenclature Committee · Database 2026
     used by CCNT2 / cyclin T2 - expressed in every KMT2A-rearranged line and needed by none of them, so the cyclin T split is real rather than an expression artefact, CDK13 - rank 171 of 17,728, equally deep in both lineage halves, and essentially unstudied in leukaemia, CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose, The transcriptional CDK with a real KMT2A-rearranged selectivity signal is CDK13, not CDK9 - and nobody has looked, What was searched, what came back empty, and what was read and not used - the P-TEFb reservoir beat's audit trail
403. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed entry P22083 (FUT4_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by CD15 (Lewis-x, SSEA-1, 3-fucosyl-N-acetyllactosamine) - a sugar, not a protein: the most-named antigen in this graph that no gene encodes, FUT4 - the myeloid alpha-1,3-fucosyltransferase, a hub gene of the KMT2A-rearranged expression module, and the gene HGNC once officially called CD15
404. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     check-named-but-unwritten.py: HGNC symbols named across the graph's entries with no entry of their own

     · Oncofusion knowledge graph (internal) 2026
     used by CD19 - the molecule this disease is treated by, named in 114 entries and written in none of them until now, CD34 - named in 132 entries of this graph and written in none, and it does not mean what the graph has been using it to mean, Four beats in a row were picked by noticing the same shape by hand; this is the instrument that finds it - 171 gene symbols named in six or more entries with no entry of their own
405. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P15391 (CD19_HUMAN), B-lymphocyte antigen CD19, 556 aa, single reviewed human record

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD19 - the molecule this disease is treated by, named in 114 entries and written in none of them until now
406. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.trial_registry
     CD19-Directed Chimeric Antigen Receptor CD19 Redirected Autologous T Cells (CART19) for Orphan Indications of Pediatric B Cell Acute Lymphoblastic Leukemia

     Grupp SA (sponsor-investigator), Children's Hospital of Philadelphia · ClinicalTrials.gov 2026
     used by CD19 CAR-T in infants with KMT2A-rearranged ALL — it works, it is manufacturable, and it does not hold
407. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     risk-lineage-switch-cd19-directed-therapy and target-cd19-car-t-infant-kmt2a-r, Oncofusion Knowledge Graph, read 2026-09-23

     · Oncofusion Knowledge Graph 2026
     used by CD19 looks lower in KMT2A-rearranged patients in the RNA-seq arm of TARGET and identical in the microarray arm of the same study - so this graph should not record a CD19 deficit
408. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 42296977
     Inotuzumab ozogamicin in paediatric very high risk first B-cell acute lymphoblastic leukaemia relapse (ITCC-059): a multicentre, single-arm, phase 2 trial.

     Peccatori N, ..., Brivio E · Lancet Haematol 2026
     used by CD22 (SIGLEC2) - the antigen that comes on BEFORE CD19, which makes it both the backup target and the pre-loaded escape route
409. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P20273 (CD22_HUMAN), B-cell receptor CD22, B-lymphocyte cell adhesion molecule, Sialic acid-binding Ig-like lectin 2, 847 aa, single reviewed human record, primary gene name CD22

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD22 (SIGLEC2) - the antigen that comes on BEFORE CD19, which makes it both the backup target and the pre-loaded escape route
410. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human entry Q9NZQ7 (PD1L1_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by CD274 (PD-L1) - named in zero entries of this graph until now, a direct MLL1-H3K4me3 target gene in another tissue, and a cell-intrinsic dependency of MLL-AF9 leukaemia-initiating cells
411. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P20138 (CD33_HUMAN), Myeloid cell surface antigen CD33, Sialic acid-binding Ig-like lectin 3, 364 aa, single reviewed human record, primary gene name CD33

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD33 (SIGLEC3) - the aberrant myeloid antigen on a B-lineage blast, and the one antibody-drug-conjugate target with a positive randomised result in KMT2A-rearranged disease
412. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P20138 (CD33_HUMAN) SUBCELLULAR LOCATION comments (isoform-scoped) and feature table, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by CD33 (SIGLEC3) - the aberrant myeloid antigen on a B-lineage blast, and the one antibody-drug-conjugate target with a positive randomised result in KMT2A-rearranged disease
413. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paperPMID 42720167
     Cuplike nuclei and bimodal CD34 expression in adult KMT2A-rearranged pro-B acute lymphoblastic leukemia.

     Xu Z, Roland K · Blood 2026
     used by CD34 - named in 132 entries of this graph and written in none, and it does not mean what the graph has been using it to mean
414. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P28906 (CD34_HUMAN), Hematopoietic progenitor cell antigen CD34, 385 aa, single reviewed human record, primary gene name CD34

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD34 - named in 132 entries of this graph and written in none, and it does not mean what the graph has been using it to mean
415. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P28907 (CD38_HUMAN), ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase 1, 300 aa, single reviewed human record, primary gene name CD38

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD38 - the one marker that is BRIGHTER in KMT2A-rearranged B-ALL, an antigen-targeted therapy that has already produced lineage switch, and an enzyme rather than a receptor
416. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record HGNC:31713 (CLEC12A)

     HUGO Gene Nomenclature Committee · HGNC 2026
     used by CD65s (VIM-2, sialylated ceramide dodecasaccharide) - the myeloid antigen that, with CD15, makes NG2 a perfect predictor of KMT2A rearrangement, CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce
417. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Europe PMC REST search for VIM-2 epitope biosynthesis, with operand controls, run 2026-09-20

     · Europe PMC 2026
     used by CD65s has no known synthesising enzyme, and the three fucosyltransferases nearest to it are explicitly excluded, Metabolic enzyme enumeration in KMT2A-rearranged leukaemia: the instrument, its positive controls, seventeen bounded absences - and the counted zero that would have been wrong, Of the 90 papers citing the discovery of KMT2A acetylation at K1130/K1133, the only three that mention MLL-rearranged disease or AFF1 are reviews, and none mentions menin - salvaged from an agent the session limit killed, The KDM6 contradiction resolved one step: GSK-J4 inhibits UTX too, so it should ANTAGONISE a menin inhibitor - and nobody has combined them
418. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P09564 (CD7_HUMAN), T-cell antigen CD7, GP40, T-cell leukemia antigen, 240 aa, single reviewed human record, primary gene name CD7

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CD7 - the T/NK antigen on a B blast that flags KMT2A rearrangement and worse event-free survival, and the fourth antigen on the lineage-switch list
419. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 5YDE and 5YDF, verified against the RCSB data API for title, experimental method and resolution

     RCSB Protein Data Bank · RCSB PDB 2026
     used by CDC73 / parafibromin — the PAF1C subunit that was genetically tested, and a second tumour suppressor required by the fusion
420. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q6P1J9 (CDC73_HUMAN), SUBCELLULAR LOCATION comment, evidence codes and feature table, fetched live 2026-09-22

     UniProt Consortium · Databases 2026
     used by CDC73 / parafibromin — the PAF1C subunit that was genetically tested, and a second tumour suppressor required by the fusion
421. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, 2026-09-21. alias-aware: the previous official symbol CDC2L5 and the alias CHED were both included, per the retired-symbol rule

     · NCBI PubMed 2026
     used by CDK13 - rank 171 of 17,728, equally deep in both lineage halves, and essentially unstudied in leukaemia
422. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap CRISPR (Chronos) gene effect, retrieved via the Open Targets Platform GraphQL API

     Broad Institute DepMap; Open Targets · Database 2026
     used by CDK9 inhibition in KMT2A-rearranged infant leukemia — the most direct hit on the hijacked machinery, and the narrowest window
423. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human entries retrieved by gene_exact:CDKN2A with fields accession, gene_primary, gene_synonym, protein_name and length

     UniProt Consortium · UniProt 2026
     used by CDKN2A - one gene, two reviewed UniProt records with the same primary symbol, and a deletion that is three times rarer here than in the infant leukaemias next door, Enumeration of candidate cooperating lesions in KMT2A-rearranged leukaemia - what was written, what was recorded as a gap, and what was read and rejected
424. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI eutils esearch counts for KMT2A/MLL disease terms intersected with candidate cell-type terms, run 2026-09-19

     Oncofusion Knowledge Graph project search · NCBI eutils 2026
     used by Cell identities considered and not written: what the search actually returned for basophils, mast cells, eosinophils, NK cells and the rest
425. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI Entrez Programming Utilities (E-utilities) esearch endpoint, db=pubmed, with querytranslation returned

     · US National Library of Medicine 2026
     used by Cell-state enumeration: 23 queries, what was written, what was rejected and why — including an empirical confirmation that PubMed reads SEM as two journal names
426. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q8N8E3 (CE112_HUMAN), Centrosomal protein of 112 kDa — accession, primary gene name, synonym, length and subcellular-location comment retrieved 2026-09-24

     · UniProt Consortium, EMBL-EBI/SIB/PIR 2026
     used by CEP112 is a 955-residue centrosomal coiled-coil protein on 7 spectra with zero leukaemia literature — its only nuclear annotation is a single 2024 citation, and a long coiled-coil is the classic non-specific AP-MS passenger — declined
427. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - Q9BUX1, CHAC1_HUMAN, glutathione-specific gamma-glutamylcyclotransferase 1

     · UniProt Knowledgebase 2026
     used by CHAC1 / BOTCH (Q9BUX1) - the largest single effect in the whole menin-inhibitor response (-5.11 log2) and a glutathione-degrading pro-apoptotic enzyme, named nowhere in this graph until now
428. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q13111 (CAF1A_HUMAN), Chromatin assembly factor 1 subunit A, 956 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CHAF1A (CAF-1 p150) - the subunit that reads PCNA and makes histone deposition replication-coupled, and the half of CAF-1 that leukaemia has not looked at
429. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q13112 (CAF1B_HUMAN), Chromatin assembly factor 1 subunit B, 559 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CHAF1B (CAF-1 p60) - the one histone chaperone in this beat with a real leukaemia literature, and a mouse experiment that contradicts DepMap's verdict on its druggability
430. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     scripts/check-pi-dimensions.py before and after the cycle-77 fix, reconciled against scripts/validate-graph.mjs type counts; this repository

     Oncofusion knowledge graph, coordinator · internal instrument audit, reproducible from the repository 2026
     used by check-pi-dimensions.py selected fusion entries by id prefix and swept in an analogy node, reporting 50/51 where the truth is 49/50 - identical to one decimal, which is why it survived - and the one genuinely unlinked fusion has no analogue in this corpus, so no edge was written, USP7 - the enzyme that holds PRC1.1 together, and the only pharmacological handle on this complex that exists today
431. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O14757 (CHK1_HUMAN), Serine/threonine-protein kinase Chk1, 476 aa, primary gene name CHEK1

     · UniProt Consortium 2026
     used by CHEK1 (CHK1) - the node where this graph's chidamide synergy actually lands, and the reason that synergy is an HR-impairment argument
432. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O96017 (CHK2_HUMAN), Serine/threonine-protein kinase Chk2, 543 aa, primary gene name CHEK2, synonyms CDS1 CHK2 RAD53

     · UniProt Consortium 2026
     used by CHEK2 (CHK2) - named in every combination mechanism here and measured on its own in none of them
433. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed esearch, db=pubmed, term '(MLL OR KMT2A) AND CHEK2', run 2026-09-20; count 2; both titles read from a full efetch record and judged off-topic

     · NCBI PubMed 2026
     used by CHEK2 (CHK2) - named in every combination mechanism here and measured on its own in none of them
434. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.news
     Children's Oncology Group — organisational overview

     Children's Oncology Group · childrensoncologygroup.org 2026
     used by Children's Oncology Group — the trial network, the AALL protocol series and the biobank that no single lab can substitute
435. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41182175
     Tandem Duplications in UBTF Create XPO1-Dependent Nuclear Export Signals that Reveal a Leukemic Therapeutic Dependency.

     Barajas JM, Phillips AH, Wang J, et al. · Blood Cancer Discovery 2026
     used by Chromatin-bound XPO1 as a shared, druggable address for HOXA-driven leukaemia - and the objection that it may be a general chromatin cofactor rather than an address at all, Docking onto pre-bound XPO1 is a general route to leukaemic chromatin, used by at least seven different lesions - not three coincidences, How a lift operator became a postal address: the idea behind the XPO1 axis, in plain language, UBTF tandem duplications INVENT an export signal, and that invented signal is what puts the protein on chromatin - the cleanest proof the XPO1 address is real
436. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     DepMap KMT2A-selective dependency scan (full genome), analyses/depmap-kmt2a-selective-scan-full

     Oncofusion Knowledge Graph project · internal analysis over DepMap CRISPR gene effect 2026
     used by Chromatin-bound XPO1 as a shared, druggable address for HOXA-driven leukaemia - and the objection that it may be a general chromatin cofactor rather than an address at all, Two reasons XPO1 is a hard target: every cell needs it, and blocking it switches on a survival pathway, XPO1 / exportin-1 (CRM1) - the cell's main nuclear export receptor, and a chromatin-bound transcriptional scaffold
437. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed E-utilities esearch, db=pubmed, run 2026-09-20

     NCBI · PubMed 2026
     used by CHUK (IKK-alpha) — the only kinase that sits in BOTH NF-kB arms, which makes it the one subunit whose inhibition has an ambiguous sign in this disease
438. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O15111 (IKKA_HUMAN), SUBCELLULAR LOCATION comment with evidence codes, sources and note, plus TISSUE SPECIFICITY, fetched live 2026-09-21

     UniProt Consortium · Databases 2026
     used by CHUK (IKK-alpha) — the only kinase that sits in BOTH NF-kB arms, which makes it the one subunit whose inhibition has an ambiguous sign in this disease
439. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human entry P33076 (C2TA_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by CIITA - the class II transactivator, a gene menin sits on, and the only antigen-presentation gene in this disease with a knockout that proves the drug acts through it
440. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41144759
     Menin inhibition enhances graft-versus-leukemia effects by T-cell activation and endogenous retrovirus induction in AML.

     Fetsch V, Schwobel L, Ozyerli-Goknar E, Stell AV, Punta M, Plenge T, et al., Zeiser R · Blood 2026
     used by CIITA - the class II transactivator, a gene menin sits on, and the only antigen-presentation gene in this disease with a knockout that proves the drug acts through it, Does menin inhibition have any MHC class II headroom to work with in B-precursor infant ALL? A flow cytometry experiment on samples that already exist, Graft-versus-leukaemia in KMT2A-rearranged infant leukaemia - the mechanism the whole transplant rests on, demonstrated in ALL generally, assumed in this disease, and with a published mechanistic reason to doubt it, Hypothesis: a menin inhibitor must come BEFORE an antigen-directed immunotherapy, not with it or after it - and every published experiment has already assumed this without anyone testing it, IFNGR1 - the receptor the whole antigen-presentation mechanism runs through, and the one where low-dose interferon-gamma makes MLL-fusion leukaemia worse, Menin inhibition restores graft-versus-leukaemia only where MHC class II is limiting - so the post-transplant maintenance rationale may hold in myeloid disease and fail in B-precursor infant ALL, MHC class I is a bivalent gene set, and a menin inhibitor derepresses it - shown in K-562, SCLC and neuroblastoma, and measured on HLA-A in primary AML, but never once in a KMT2A-rearranged cell, Raising MHC class I is not a free good: the same molecule is the inhibitory ligand that holds NK cells off, and in this fusion family it is an in vivo leukaemia dependency, The class I and class II arms of menin inhibition are different mechanisms - the class I effect was tested against viral mimicry and NF-kB and excluded both, The interferon state is not baseline in this disease — it is something a menin inhibitor creates, through endogenous retrovirus induction and MHC class II
441. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q03164 (KMT2A_HUMAN), Modified residue feature table (ft_mod_res)

     · UniProt Consortium 2026
     used by CK2 phosphorylation of KMT2A S136/S142 tunes the LEDGF grip - a second, non-menin, phospho-gated handle on the fragment every fusion keeps
442. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.preprint
     Menin-Inhibition Sensitizes Acute Myeloid Leukemia to CLEC12A-Directed CAR Cell Therapy

     Rausch J, Wendel P, Dzama-Karels M, Steiner M, Gierschek F, Fetsch V, et al. · Preprint, not peer reviewed (Europe PMC PPR1223455) 2026
     used by CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce, Hypothesis: a menin inhibitor must come BEFORE an antigen-directed immunotherapy, not with it or after it - and every published experiment has already assumed this without anyone testing it, Menin inhibition reported to induce CLEC12A and sensitise KMT2A-r and NPM1-mutant AML to CLEC12A-directed CAR cells (2026 preprint, not peer reviewed), The NK cell - the effector arm of haploidentical transplant, the one this disease is documented to evade, and the cell CLEC12A was named on
443. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41941266
     Phase 1 Study of KITE-222, an Autologous CLL-1-Directed CAR T-cell Therapy in Patients with Relapsed/Refractory Acute Myeloid Leukemia.

     Daver N, Blachly JS, Ghobadi A, Advani A, Muffly L, Garciaz S, et al. · Clin Cancer Res 2026
     used by CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce, Clec12a is REQUIRED for NUP98::NSD1 AML pathogenesis - a marker turning out to be a dependency, in a different fusion family, and untested in KMT2A-r, Experiment: CRISPR knockout of CLEC12A in KMT2A-rearranged models, with and without a cooperating RAS lesion, Hypothesis: CLEC12A is a cell-intrinsic dependency in KMT2A-rearranged leukaemia, not only a surface marker - and if so, antigen-negative escape is self-defeating, Menin inhibition reported to induce CLEC12A and sensitise KMT2A-r and NPM1-mutant AML to CLEC12A-directed CAR cells (2026 preprint, not peer reviewed), Q48 enumeration: every antigen, identifier and query this beat ran - including what was searched and rejected, and the three identifier traps found live, Risk: CLEC12A-directed therapy ablates normal granulocytes and disables an inhibitory brake on NET formation - the toxicity is mature-myeloid, not stem-cell, The NK cell - the effector arm of haploidentical transplant, the one this disease is documented to evade, and the cell CLEC12A was named on, Which immune-evasion routes are actually documented in KMT2A-rearranged disease, and which are borrowed from AML and assumed - a ranked ledger with the bounded absences that define it
444. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed entry Q5QGZ9 (CL12A_HUMAN): sequence length, topology, domain and motif feature table, and Pfam/InterPro/PROSITE cross-references, fetched live from the UniProt REST API

     UniProt Consortium · UniProtKB 2026
     used by CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce
445. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.trial_registry
     ClinicalTrials.gov sponsor records for Biomea Fusion, retrieved via the ClinicalTrials.gov API v2

     US National Library of Medicine; Biomea Fusion · ClinicalTrials.gov 2026
     used by Collaborate or compete: an honest matrix, including where the lab would be duplicating a better-resourced effort
446. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed E-utilities esearch, db=pubmed, run 2026-09-20

     NCBI · PubMed 2026
     used by Combine an IKK inhibitor with a clinical menin inhibitor in a KMT2A::AFF1 B-ALL line, and read out fusion OCCUPANCY rather than viability, IKK inhibition and menin inhibition evict the fusion from chromatin by routes that do not share a node, so they should be additive — and the same reasoning predicts that a menin inhibitor alone partially defeats itself. Prior 0.30
447. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/does-the-stored-author-match-the-doi/run.py, results.json and author-index.json, this repository, computed 2026-09-24 with author lists fetched live from the Crossref REST API

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Comparing every citation's stored author list against the paper its DOI resolves to - 15,219 comparable blocks, 99.6% agreeing, and all 60 disagreements are placeholder strings, one Crossref metadata error and one Scandinavian letter - which closes the year check's known blind spot and finds no wrong-paper citation on either axis
448. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/does-the-stored-year-match-the-doi/run.py and results.json, this repository, computed 2026-09-24 against the committed node files and the resolved Crossref records cached in analyses/do-this-corpus-s-dois-resolve/doi-index.json

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Comparing every citation's stored year and journal against the record its DOI actually points at - 16,420 comparable blocks, 92.0% exact, and all 22 that disagree by two years or more resolve on reading to six DOIs with online-versus-print date conventions, so the wrong-paper failure mode is bounded from above for the first time
449. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41576348
     Advancing precision therapy in pediatric acute myeloid leukemia through PDX models and mitochondrial targeting.

     Da Ros A, Peloso A, Longo G, ..., Pigazzi M · Blood Adv 2026
     used by Complex I inhibition triggers the mitochondrial integrated stress response through DELE1 and HRI, and that is what creates the MCL-1 dependency - a metabolic drug manufacturing an apoptotic vulnerability, Mitochondrial complex I plus venetoclax in KMT2A-rearranged paediatric AML - a genotype-stratified in vivo PDX result, on a drug whose adult clinical development stopped for toxicity, The experiment that separates lineage from ontogeny: one KMT2A fusion, two cell-of-origin contexts, three metabolic perturbations, and a limiting-dilution readout alongside the viability one
450. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     AlphaFold Protein Structure Database prediction API, Q03164

     · EMBL-EBI 2026
     used by Compute: which backend for which piece of phase-2 work, and the one constraint that matters for each
451. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42285105
     RNA reinforces condensate nucleation on chromatin to amplify oncogenic transcription

     Budinich KA, Yao X, Gong C, et al. (Wan L, senior author) · Molecular Cell 86(14):2777-2793.e12 2026
     used by Condensates in KMT2A-fusion leukemia — an honest verdict: not established for AFF1/AFF4/AF9, load-bearing only in ENL-mutant and NUP98 fusions
452. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     Reading-depth audit: self-declared abstract-only citation across the Oncofusion knowledge graph

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Confidence in this graph carries no information about whether anyone read the paper
453. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42001795
     Characterizing the impact of MLL fusion variants and fusion partners on leukemia plasticity using a human CRISPR-engineered MLL-rearranged leukemia model

     · Neoplasia 2026
     used by Cord blood does transform — when the translocation is engineered rather than expressed, Lineage switch in KMT2A-rearranged leukemia — the same clone changes what kind of blood cell it is, Re-bound at full text: 100 records pair an MLLT1 alias with reciprocal-fusion language and 8 with spliced-fusion language, and none stratifies a KMT2A::MLLT1 outcome, transcriptome or drug response by reciprocal or spliced status
454. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Cellosaurus CVCL_0093 (RS4;11): Sequence variation - Gene fusion AFF1 + KMT2A (KMT2A-AFF1, MLL-AF4), in frame

     · Cellosaurus (SIB) 2026
     used by CORRECTS A DISPATCH PREMISE: GSE149183's menin-versus-DOT1L-inhibitor head-to-head exists in MOLM13 (KMT2A::MLLT3) ONLY. RS4;11 - the index fusion - is in the series but carries H3K79me2 alone, with no VTP-50469 arm, no MEN1 ChIP and no DOT1L ChIP, and the two arms are on different genome builds, RS4;11 (CVCL_0093) — an ADULT KMT2A::AFF1 line, MSI-high, carrying NSD2 E1099K, The PPTC childhood PDX panel — 90 leukaemia models, an MLL-ALL group, and no published partner or age breakdown
455. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, db=pubmed: four AFF1/AF4/MLLT2/FEL alias queries crossed with phase-separation terms, run 2026-09-20

     Not applicable - database query · NCBI E-utilities 2026
     used by Counted absence: AFF1/AF4 - the partner in the commonest infant fusion - has no phase-separation literature under any alias, and the two hits that look like it are token collisions
456. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     REVUFORJ (revumenib) US prescribing information, indications and usage

     Syndax Pharmaceuticals · FDA Structured Product Label via openFDA, SPL id 76183831-b710-4be2-a6a7-9349dd501830, effective 2026-02-04 2026
     used by Counted absence: no menin-inhibitor trial has ever reported a response RATE in NUP98-rearranged patients
457. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     T2023-004 Phase 2 Trial of Revumenib (SNDX-5613) in Combination With FLA Chemotherapy for Children With Relapsed or Refractory NUP98-rearranged Acute Myeloid Leukemia

     Therapeutic Advances in Childhood Leukemia Consortium (sponsor) · ClinicalTrials.gov, record last update posted 2026-07-22 2026
     used by Counted absence: no menin-inhibitor trial has ever reported a response RATE in NUP98-rearranged patients
458. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     Comparing symbol-keyed and complex-phrase PubMed queries for every complex this graph models, to find counted absences that are artefacts of the key

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Counted absences keyed on a complex subunit miss the papers that name only the complex - demonstrated on our own SSRP1 entry, which declared no KMT2A paper about FACT while this graph cites one about the index fusion in four other entries
459. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     counter-note-boilerplate-census/run.py: near-duplicate clustering of every counter_evidence note in the graph

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Counter-argument reuse across 8,080 notes is 1.04x and only 1.6% are class-level objections that ignore their own claim: the ledger's standing quality worry was measured on 53 notes and does not generalise
460. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     counter-note-boilerplate-census/class_level.py: class-level template detection cross-tabulated against engagement with the note's own claim

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Counter-argument reuse across 8,080 notes is 1.04x and only 1.6% are class-level objections that ignore their own claim: the ledger's standing quality worry was measured on 53 notes and does not generalise
461. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     Ranking curated interactors of the KMT2A::AFF1 scaffold and its 3' partners by distinct laboratory rather than by publication count

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Counting laboratories instead of papers demoted nothing, because BioPlex 2.0 and 3.0 have different last authors - and the 3' partners' own neighbourhoods are the emptiest part of the graph
462. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed records for MLLT1 (Q03111), MLLT3 (P42568), MLLT10 (P55197), AFF3 (P51826), ELL (P55199) and ELL2 (O00472), fetched with gene_exact and the primary gene name read back

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by Counting laboratories instead of papers demoted nothing, because BioPlex 2.0 and 3.0 have different last authors - and the 3' partners' own neighbourhoods are the emptiest part of the graph, Of 262 curated interactors this graph never wrote down, five are selectively required in KMT2A-rearranged lines - and one of them is FANCD2, which binds menin and sits inside a selectivity signal this project has failed four times to explain
463. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     Register of on-topic literature known to exist and not retrievable by this project's APIs

     · Oncofusion knowledge graph (internal register) 2026
     used by Covalent or reversible is the axis on which ‘the menin inhibitor class’ most plausibly breaks
464. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P50416 (CPT1A_HUMAN) - Carnitine O-palmitoyltransferase 1, liver isoform (CPT1-L); also annotated Succinyltransferase CPT1A

     UniProt Consortium · UniProtKB 2026
     used by CPT1A - the outer-mitochondrial-membrane fatty-acid gatekeeper that is also a lysine succinyltransferase, fused to KMT2A exon 3 in the one case that is tether WITHOUT address
465. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     RefSeq NM_001876.4 - Homo sapiens carnitine palmitoyltransferase 1A (CPT1A), transcript variant 1, mRNA

     NCBI RefSeq · NCBI Nucleotide 2026
     used by CPT1A - the outer-mitochondrial-membrane fatty-acid gatekeeper that is also a lysine succinyltransferase, fused to KMT2A exon 3 in the one case that is tether WITHOUT address
466. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset
     DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan)

     Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
     used by CPT1A - the outer-mitochondrial-membrane fatty-acid gatekeeper that is also a lysine succinyltransferase, fused to KMT2A exon 3 in the one case that is tether WITHOUT address
467. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 42378254
     Phosphorylated DEK sustains leukemia stem cells by enabling PBX3-driven transcriptional reprogramming.

     Lei Y, et al. · Blood 2026
     used by CSNK2A1 (casein kinase 2 alpha) - the one kinase that turns up three separate times in KMT2A biology, DEK - a chromatin architectural protein that gatekeeps H3.3 and boosts HP1alpha, is a fusion partner in its own right, and is the substrate behind this graph's third CK2 route, Separating CK2's two routes into KMT2A leukaemia, which every published experiment has confounded
468. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC records for CSNK2A1 (HGNC:2457, P68400) and CSNK2A2 (HGNC:2459, P19784), the latter carrying CSNK2A1 as an alias symbol

     · HUGO Gene Nomenclature Committee 2026
     used by CSNK2A1 (casein kinase 2 alpha) - the one kinase that turns up three separate times in KMT2A biology
469. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.preprintPMID 42039464
     Lineage-specific CK2α deletion reshapes the transcriptome of hematopoietic stem cells toward an immune-primed state.

     Valensi H, Rajaiah R, Shanmugam M, Muhammad D, Golla U, Mercer K, Karampuri A, Dovat S, Behura C, Uzun Y · bioRxiv 2026
     used by CSNK2A1 (casein kinase 2 alpha) - the one kinase that turns up three separate times in KMT2A biology
470. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     gene-csnk2a1 in this knowledge graph: CSNK2A1 (P68400, 391 aa) reaches KMT2A biology through three independent routes - LEDGF-binding motif phosphorylation at KMT2A S136/S142 on the retained fragment, taspase1-adjacent phosphorylation at T2724/S2726 on the lost fragment, and DEK phosphorylation in leukaemia stem cells

     Oncofusion Knowledge Graph, Q22 phospho-contacts beat · Internal knowledge graph entry 2026
     used by CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose
471. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P19784 (CSK22_HUMAN) and O54833 (CSK22_MOUSE), SUBCELLULAR LOCATION comments with evidence codes, sources and notes, both fetched live 2026-09-21

     UniProt Consortium · Databases 2026
     used by CSNK2A2 - CK2's other catalytic subunit, expressed in every KMT2A-rearranged line and, if anything, slightly favourable to lose
472. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     PubMed E-utilities esearch, db=pubmed, run 2026-09-20: three queries intersecting NG2/CSPG4 with menin-inhibitor terms, each returning count 0, with querytranslation inspected

     · NCBI PubMed 2026
     used by CSPG4 / NG2 - the flow-cytometry surrogate for a broken KMT2A, and a direct target gene of the fusion that makes it, NG2 reads the fusion's transcriptional programme, not the DNA lesion - so a menin inhibitor should erase the graph's best diagnostic marker while the rearrangement stays put
473. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q6UVK1 (CSPG4_HUMAN), Chondroitin sulfate proteoglycan 4, 2322 aa, single reviewed human record, primary gene name CSPG4

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CSPG4 / NG2 - the flow-cytometry surrogate for a broken KMT2A, and a direct target gene of the fusion that makes it
474. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42345361
     NG2-ITGA4 axis regulates Rho GTPases and leukemic aggressiveness in KMT2A-r B-ALL and is targetable with natalizumab.

     Rubio-Gayarre A, Vinyoles M, Tejedor JR, Guerrero-Murillo M, Fernandez-Fuentes N, Caracuel-Peramos R, Plaza-Calonge MC, Martinez-Moreno A, Roca-Ho H, Petazzi P, Pañella-Alférez A, Meseguer-Giron A, Vrenken KS, Fraga MF, Leoz Allegretti P, Palacio-Garcia C, Velasco P, Fuster JL, Stam RW, Rodríguez-Manzaneque JC, Bustelo XR, Robles-Valero J, Menendez P, Bueno C, Lopez-Millan B · Blood 2026
     used by CSPG4 / NG2 - the flow-cytometry surrogate for a broken KMT2A, and a direct target gene of the fusion that makes it, ITGA4 (CD49d, the alpha-4 of VLA-4) - the niche adhesion subunit with an approved antibody and a 2026 KMT2A-r B-ALL xenograft result, ITGA4 (VLA-4) and NG2 — the niche-adhesion axis of KMT2A-rearranged B-ALL, and the one approved drug that hits it, NG2+ periarteriolar pericyte — the cell that holds blood stem cells asleep, and the one MLL-AF9 leukaemia destroys
475. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed entry Q6UVK1 (CSPG4_HUMAN), Chondroitin sulfate proteoglycan 4 - accession, length, topological domains, domains and Pfam cross-references

     · UniProt 2026
     used by CSPG4 / NG2 - the flow-cytometry surrogate for a broken KMT2A, and a direct target gene of the fusion that makes it
476. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P49711 (CTCF_HUMAN), Transcriptional repressor CTCF, 727 aa, primary gene name CTCF

     · UniProt Consortium 2026
     used by CTCF - named in the id of an existing entry in this graph and having no entry of its own, and the protein whose binding sites break preferentially in KMT2A-rearranged patients
477. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB search gene_exact:CTH AND organism_id:9606 AND reviewed:true, and entry P32929 (CGL_HUMAN), Cystathionine gamma-lyase

     UniProt Consortium · UniProt 2026
     used by CTH (P32929) - gene_exact returns TWO reviewed records and the second is VSIG2, a third instance of rule 62 on the same symbol the Ensembl resolver already got wrong
478. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI E-utilities esearch against PubMed, 2026-09-22, fusion operand set (KMT2A[tiab] OR MLL[tiab] OR "MLL-AF4"[tiab] OR "KMT2A::AFF1"[tiab] OR "MLL-rearranged"[tiab] OR "t(4;11)"[tiab] OR "MLL-AF9"[tiab] OR "MLL-ENL"[tiab]), each term separately tagged and every cytogenetic term quoted

     · PubMed 2026
     used by CTH (P32929) - gene_exact returns TWO reviewed records and the second is VSIG2, a third instance of rule 62 on the same symbol the Ensembl resolver already got wrong
479. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P35222 (CTNB1_HUMAN) SUBCELLULAR LOCATION comments and note, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by CTNNB1 (beta-catenin) — the Wnt link inside DotCom, and the one piece of KMT2A biology that is not chromatin
480. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed search for CTR9 in the context of KMT2A/MLL-fusion leukemia, conducted for this graph on 2026-09-18; no primary subunit-level study retrieved

     Oncofusion Knowledge Graph project, using the NCBI PubMed E-utilities · Literature search 2026
     used by CTR9 — a PAF1C subunit the field assumes is a fusion partner, with no paper that tested it
481. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB Q6PD62 (CTR9_HUMAN) feature table and PDB cross-references; RCSB entries 6TED, 7OPC and 8A3Y verified via the RCSB data API

     UniProt Consortium; RCSB Protein Data Bank · Databases 2026
     used by CTR9 — a PAF1C subunit the field assumes is a fusion partner, with no paper that tested it
482. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q6PD62 (CTR9_HUMAN) and Q62018 (CTR9_MOUSE), SUBCELLULAR LOCATION comments with evidence codes and sources, both fetched live 2026-09-21

     UniProt Consortium · Databases 2026
     used by CTR9 — a PAF1C subunit the field assumes is a fusion partner, with no paper that tested it
483. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/q122-h3k79me2-three-arms-with-replicates/results.json, criteria C1, C2, C3, C6, C8, R1, R3 and R5; pre-registration written before any bin file existed; this repository

     Oncofusion knowledge graph, Q122 beat (computation) and coordinator (curation of the numbers into this entry) · internal analysis, reproducible from the repository 2026
     used by CURATED from committed-but-unread results: murine MLL-AF9 (MLLT3) retains H3K79me2 like human MOLM13 (MLLT3) and unlike RS4;11 (AFF1) - four confirmations fire, two refutations do not, the cross-partner delta is 3.2-3.6x the between-culture spread, and the flatness survives an amplitude stratification in 8 of 8 measurements, CURATED: Q120 called the MLL1-N channel 'inert to three decimals'. At MLL1's OWN top sites in mouse it falls to 0.3012 under a DOT1L inhibitor and 0.3298 under a menin inhibitor - so 'inert' was population-specific, and a drug that does not touch menin moves the MLL1 channel as much as one that does, SALVAGED from a beat killed mid-write: with direct DOT1L inhibition as the positive control that bounds the achievable effect, VTP-50469 achieves only F = 0.09-0.25 of it. SGC0946 strips H3K79me2 to 6.7-14.8% of DMSO; the menin inhibitor leaves 77-91%. And the bin-wise responses do not correlate (rho = -0.026), SALVAGED: in K562, which carries no KMT2A fusion at all, a menin inhibitor moves H3K79me2 as much as it does in fusion-bearing murine MLL-AF9 (V 0.8291 vs 0.7734, F 0.1831 vs 0.2500). Both the confirmation and the refutation fail to fire, and the refutation misses by 0.0057 - this is at-the-edge and must not be read as a clean negative
484. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P48061 (SDF1_HUMAN) - Stromal cell-derived factor 1

     · UniProtKB 2026
     used by CXCL12 (SDF-1) - the marrow retention chemokine this graph already had a mechanism entry for and no entry for the protein
485. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P61073 (CXCR4_HUMAN) - C-X-C chemokine receptor type 4

     · UniProtKB 2026
     used by CXCR4 - named in eighteen entries of this graph with no entry of its own, and the one niche molecule with a drug, a paediatric trial and a direct KMT2A-r experiment
486. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch, db=pubmed - bounded absence query with querytranslation readback

     · NCBI PubMed 2026
     used by CXCR4 - named in eighteen entries of this graph with no entry of its own, and the one niche molecule with a drug, a paediatric trial and a direct KMT2A-r experiment, Nobody has reported what a menin inhibitor does to CXCR4: zero papers name menin and CXCR4 together, and the RNA-seq that would answer it is already public, Planned: read CXCR4 and SYTL1 out of the menin-inhibitor transcriptomes that are already public, in the same contrast that shows HOXA9 and MEIS1 falling
487. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     GSE281506 - Combining Menin and MEK inhibition to target poor prognosis KMT2A-rearranged RAS pathway-mutant acute myeloid leukemia (expression profiling by high throughput sequencing, Homo sapiens, 42 samples)

     · NCBI Gene Expression Omnibus 2026
     used by CXCR4 function in KMT2A-rearranged blasts is an output of the fusion's own HOXA/MEIS1 program, so a menin inhibitor should switch off the niche axis without a CXCR4 antagonist, Nobody has reported what a menin inhibitor does to CXCR4: zero papers name menin and CXCR4 together, and the RNA-seq that would answer it is already public, Planned: read CXCR4 and SYTL1 out of the menin-inhibitor transcriptomes that are already public, in the same contrast that shows HOXA9 and MEIS1 falling
488. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9P0U4 (CXXC1_HUMAN), CXXC-type zinc finger protein 1

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by CXXC1 (CFP1) - a CpG-reading COMPASS subunit that binds MLL1 and recruits it to HoxA7, and whose CXXC domain cannot substitute for KMT2A's in a fusion
489. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch record counts for CXXC1/CFP1 against KMT2A, MLL and leukaemia, run for interactome round 5

     · NCBI PubMed, queried 2026-09-19 2026
     used by CXXC1 (CFP1) - a CpG-reading COMPASS subunit that binds MLL1 and recruits it to HoxA7, and whose CXXC domain cannot substitute for KMT2A's in a fusion
490. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9H2H0 (CXXC4_HUMAN), CXXC-type zinc finger protein 4 / Inhibition of the Dvl and axin complex protein (IDAX)

     UniProt Consortium · UniProtKB 2026
     used by CXXC4 (IDAX) - TET2's CXXC domain, sitting on a separate gene after an evolutionary inversion, and annotated cytoplasmic only by similarity
491. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q7LFL8 (CXXC5_HUMAN), CXXC-type zinc finger protein 5 / Retinoid-inducible nuclear factor (RINF)

     UniProt Consortium · UniProtKB 2026
     used by CXXC5 (RINF) - the one CXXC reader measurably down-regulated in KMT2A-rearranged AML, which is the only direction-of-change any family member has in this disease
492. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     analyses/count-threshold-transferability/run.py and results.json, this repository; run 2026-09-23 against the working tree at 2,301 node files

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Cycle 76 warned that count thresholds in falsifiers may not transfer across set sizes. Sweeping all 2,301 entries found 5 scored criteria with a count threshold and exactly ONE that is both fragile and live - hyp-q117's uncashed PEAK ARM, now amended before it runs
493. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/epigenetics-field-versus-prose/run.py and results.json, this repository; run 2026-09-23, sample seed 20260923

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Cycle 77 published epigenetics as this graph's weakest PI dimension at 35.9% by dedicated field. Reading the entries that have no field shows 10 of 10 carry genuine, specific, cited epigenetic annotation in prose - so the dimension is 56.2% (at least 50.4% on a Wilson bound) and about 20 points of the 'gap' was a formatting artefact
494. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
     analyses/depmap-arid1b-lineage-conditioned/run.py and results.json, this repository; DepMap Chronos via Open Targets GraphQL, ARID1A variant annotation from Cellosaurus, pulled 2026-09-23

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Cycle 81 named lineage as the single biggest hole in the ARID1B paralogue-buffering result. Permuting ARID1A status only WITHIN lineage strata moves the estimate from -0.2985 to -0.2961 - a change of 0.0024 - with p = 0.00140, the instrument check surviving, both specificity controls flat and zero mutant lines dropped. Lineage explains none of it
495. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/depmap-arid1b-truncating-vs-missense/run.py and results.json, this repository; mutations from cBioPortal ccle_broad_2025, gene effects from DepMap Chronos via Open Targets GraphQL (dataVersion 26.06), pulled 2026-09-23

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by Cycle 81's pre-registered PRIMARY stratum was underpowered at n=4 for five cycles. A different source gives n=124: ARID1B buffering is -0.2475 (p=5e-05) in truncating-ARID1A lines and +0.0117 (p=0.66) in missense-only ones - a complete dose-response. My own specificity control failed, and it failed because I chose it from inside the complex I was studying
496. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, db=pubmed: (MLL OR KMT2A) AND (GPHN OR gephyrin), run 2026-09-20

     Not applicable - database query · NCBI E-utilities 2026
     used by Cytoplasmic localisation in KMT2A fusions: the SEPTIN case has an image, the GPHN case has none, and neither partner wins the compartment
497. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     cBioPortal public REST API: /api/studies, /api/structural-variant/fetch (molecularProfileIds aml_stjude_2024_structural_variants, entrezGeneIds 4297)

     · cBioPortal for Cancer Genomics 2026
     used by Dataset: cBioPortal - open, easy, and containing exactly two KMT2A::AFF1 fusion records
498. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL REST API, target search and activity endpoints with pchembl_value >= 6

     · EMBL-EBI ChEMBL 2026
     used by Dataset: ChEMBL chemical matter - open, quantitative, and easy to misread as a druggability score, Phase-2 dossier: the menin MBM1 pocket - the only candidate here that is already de-risked, and therefore the only one whose gate is clinical
499. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     DepMap Forum: Provide an open endpoint for latest version retrieval (DepMap staff response on the captcha layer)

     · DepMap Forum 2026
     used by Dataset: DepMap CRISPR essentiality - the bulk download is now behind a captcha, and this changes how phase 2 must be built
500. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     gnomAD GraphQL API, gnomad_constraint for eight candidate genes (GRCh38)

     · Broad Institute 2026
     used by Dataset: gnomAD constraint - an open, instant tolerability prior that cannot discriminate between our candidates, ENL is the highest-expressed target in cerebellum and nobody has studied it there, The phase-2 go/no-go register: the numbers we agree to before we look
501. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     GTEx Portal API v2, reference/gene and expression/medianGeneExpression endpoints

     · GTEx Consortium 2026
     used by Dataset: GTEx normal-tissue expression - fully open, and the version-string trap that silently returns nothing, Phase-2 dossier: the menin MBM1 pocket - the only candidate here that is already de-risked, and therefore the only one whose gate is clinical
502. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     dbGaP study phs000424 (Genotype-Tissue Expression Project), Authorized Access requirements

     · NCBI dbGaP 2026
     used by Dataset: GTEx normal-tissue expression - fully open, and the version-string trap that silently returns nothing
503. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Human Protein Atlas gene record, MEN1 (ENSG00000133895.json)

     · Human Protein Atlas 2026
     used by Dataset: Human Protein Atlas - the protein-level and single-cell check that GTEx cannot give, MEN1 / menin — the adaptor the fusion cannot work without, The tissue selectivity of MEN1 syndrome is not in MEN1: menin is expressed evenly everywhere and MLL, p27 and p18 are the endocrine-predominant ones
504. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB data API, core entry records for 4GQ6, 7UJ4, 3U88, 2MSR, 4HRA, 5J9S, 6HT1, 6MAJ, 3I91, 5EQ0, 2J2S, 4NW3

     · RCSB PDB 2026
     used by Dataset: PDB and AlphaFold - the structural census, including the two absences that shape every plan, Phase-2 dossier: the fusion junction - a negative feasibility finding, with the one surviving modality named, Phase-2 dossier: the menin MBM1 pocket - the only candidate here that is already de-risked, and therefore the only one whose gate is clinical
505. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     AlphaFold Protein Structure Database prediction API, queried for O00255, O95251, Q03111, O75475, Q9HC52, Q8TEK3, P51825 and Q03164

     · EMBL-EBI 2026
     used by Dataset: PDB and AlphaFold - the structural census, including the two absences that shape every plan, Phase-2 dossier: the menin MBM1 pocket - the only candidate here that is already de-risked, and therefore the only one whose gate is clinical
506. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     St. Jude Cloud documentation: Genomics Platform overview and Data Sets / Data Access Units

     · St. Jude Children's Research Hospital 2026
     used by Dataset: St. Jude Cloud - the right cohort for this disease, behind a Data Access Agreement we have not applied for
507. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCI Genomic Data Commons API, projects and files endpoints filtered to program TARGET; dbGaP study pages phs000464 and phs000465

     · NCI GDC / NCBI dbGaP 2026
     used by Dataset: TARGET via the GDC - open expression and mutations, but the fusion calls themselves are controlled
508. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.analysis
     Triaging the 262 curated interactors this graph has not written by DepMap selectivity in KMT2A-rearranged cell lines

     · Oncofusion knowledge graph (internal analysis) 2026
     used by DAXX - the strongest of five selective dependencies among 254 unwritten interactors but it FAILS Bonferroni correction (p=0.0128, rank 4 of 254), and not one paper connects it to this disease, HDGFL2 / HRP-2 - the second IBD protein, which binds KMT2A at the same site with the same affinity and does NOT need menin to do it
509. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9UER7 (DAXX_HUMAN), Death domain-associated protein 6, 740 aa

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by DAXX - the strongest of five selective dependencies among 254 unwritten interactors but it FAILS Bonferroni correction (p=0.0128, rank 4 of 254), and not one paper connects it to this disease
510. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ChEMBL target and activity records for DAXX (CHEMBL5724773), ATRX (CHEMBL4879474), menin (CHEMBL1615381) and the menin-MLL interaction (CHEMBL2093861)

     · ChEMBL, EMBL-EBI 2026
     used by DAXX and ATRX as drug targets: six ChEMBL rows and zero of them an affinity, against menin's 1,776 - but three crystal structures of the interface that would have to be blocked
511. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     Bounded PubMed absence for DAXX degraders and small molecules, run through scripts/pubmed_safe.py with operand controls

     · PubMed (NCBI E-utilities), via this project's pubmed_safe wrapper 2026
     used by DAXX and ATRX as drug targets: six ChEMBL rows and zero of them an affinity, against menin's 1,776 - but three crystal structures of the interface that would have to be blocked
512. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - P35638, DDIT3_HUMAN, DNA damage-inducible transcript 3 protein

     · UniProt Knowledgebase 2026
     used by DDIT3 / CHOP (P35638) - the cascade's middle node, and the symbol whose two reviewed human records and 16,893-record alias are the worst identifier trap in this beat
513. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB - P0DPQ6, DT3UO_HUMAN, DDIT3 upstream open reading frame protein

     · UniProt Knowledgebase 2026
     used by DDIT3 / CHOP (P35638) - the cascade's middle node, and the symbol whose two reviewed human records and 16,893-record alias are the worst identifier trap in this beat
514. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q13206, Probable ATP-dependent RNA helicase DDX10 (DDX10)

     UniProt Consortium · UniProtKB 2026
     used by DDX10 domain architecture (UniProt Q13206), NUP98::DDX10, inv(11)(p15q22) — the 'helicase' fusion whose helicase activity turns out not to be the point
515. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database (EMBL-EBI), PSICQUIC MITAB 2.7, pubid:39251607, retrieved 2026-09-23

     2026
     used by DDX55 - a large-subunit rRNA helicase that is nucleolar by imaging and nucleoplasmic by its own founding paper, and the narrowest RBP bait to reach KMT2A after DDX52
516. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     Human Protein Atlas v25.1, proteinatlas.json.gz record for DDX55 (ENSG00000111364), retrieved 2026-09-23

     2026
     used by DDX55 - a large-subunit rRNA helicase that is nucleolar by imaging and nucleoplasmic by its own founding paper, and the narrowest RBP bait to reach KMT2A after DDX52
517. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI PubMed esearch via scripts/pubmed_safe.py, run 2026-09-23; querytranslations recorded: "DDX55"[Title/Abstract] AND "KMT2A"[Title/Abstract] and "DDX55"[Title/Abstract] AND "MLL"[Title/Abstract]

     2026
     used by DDX55 - a large-subunit rRNA helicase that is nucleolar by imaging and nucleoplasmic by its own founding paper, and the narrowest RBP bait to reach KMT2A after DDX52
518. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P26196 (DDX6_HUMAN), SUBCELLULAR LOCATION comment with all evidence codes, sources and notes, plus TISSUE SPECIFICITY, fetched live 2026-09-21

     UniProt Consortium · Databases 2026
     used by DDX6 - the only protein ever proposed to do for a cell what Tat does for the virus, on the strength of one 2016 paper
519. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
     A Phase I-II Study Investigating the All Oral Combination of the Menin Inhibitor SNDX-5613 With Decitabine/Cedazuridine (ASTX727) and Venetoclax in Acute Myeloid Leukemia (SAVE)

     M.D. Anderson Cancer Center · ClinicalTrials.gov 2026
     used by Decitabine - degrades DNMT1, and in ALL cells it degrades KMT2A too: the hypomethylating agent with two mechanisms and a menin-inhibitor combination already in the clinic
520. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     ChEMBL molecule record CHEMBL1201129 (DECITABINE)

     ChEMBL, EMBL-EBI · ChEMBL database 2026
     used by Decitabine - degrades DNMT1, and in ALL cells it degrades KMT2A too: the hypomethylating agent with two mechanisms and a menin-inhibitor combination already in the clinic
521. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P48735 (IDHP_HUMAN) - retrieved alongside the UniProt CXXC-type zinc-finger family queries used to confirm family membership

     UniProt Consortium · UniProtKB 2026
     used by Decitabine degrades DNMT1 - and in ALL cells the same treatment lowers KMT2A protein, with the shared CXXC-family DNA-binding module offered as the reason, IDH2 - the hypermethylating co-mutation that IS enriched in KMT2A-rearranged AML (16.0% vs 10.4%), at p = 0.0004 and not the p < 0.0001 the abstract claims
522. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42096476
     The KMT2F histone methyltransferase interacts with the RNA polymerase I machinery to promote ribosomal RNA transcription.

     Lone KA, Karole AM, et al. · PLoS Biol 2026
     used by Degrade one CXXC reader, ChIP the fusion: the one experiment that settles whether the address is contested, Three CXXC-pathway activities converge on rDNA - KDM2B represses Pol I, KMT2A and SETD1A bind the whole locus, MLL-AF4 is predominantly nucleolar - and no paper contains two of them
523. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P35659 (DEK_HUMAN), Protein DEK - FUNCTION, SUBUNIT, SUBCELLULAR LOCATION, TISSUE SPECIFICITY and feature table, fetched 2026-09-22

     · UniProt Consortium 2026
     used by DEK - a chromatin architectural protein that gatekeeps H3.3 and boosts HP1alpha, is a fusion partner in its own right, and is the substrate behind this graph's third CK2 route
524. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
     analyses/depmap-arid1b-kmt2a-rearranged/run.py and results.json, this repository; DepMap Chronos via Open Targets GraphQL, KMT2A-fusion annotation from Cellosaurus, pulled 2026-09-23

     Oncofusion knowledge graph, coordinator · internal analysis, reproducible from the repository 2026
     used by DepMap agrees with the mouse genetics: ARID1B is not a FUSION-SELECTIVE dependency (+0.0145, p=0.80) and reads -0.08 in both groups - but CYCLE 81 SHOWED THAT FLATNESS IS PARALOGUE BUFFERING, not dispensability. ARID1A and ARID2 ARE dependencies and are equally so in both groups - so no BAF subunit tested is fusion-selective. And MEN1 (-1.26) is a harder and more selective dependency than KMT2A itself (-0.81)
525. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Cellosaurus cell-line search API, queries var:NUP98 AND dr:DepMap and var:<PARTNER> AND dr:DepMap over 27 symbols

     SIB Swiss Institute of Bioinformatics · Cellosaurus 2026
     used by DepMap cannot answer whether MEN1 is selectively required in NUP98-rearranged lines, because it contains none — and it takes only ~3 to answer it
526. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.analysis
     Diffing the curated interactome of the KMT2A::AFF1 core scaffold against the graph's own per-protein entries

     · Oncofusion knowledge graph (internal analysis) 2026
     used by Diffing the core scaffold's curated interactome against the graph: 46 of 217 partners are things here, CDK9 is not one of them, and the filter I built to rank the rest was counting one platform twice
527. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     Proteins API proteomics-PTM, O00255 (MEN1_HUMAN): 4 phosphorylation sites

     · EMBL-EBI 2026
     used by DNA damage raises menin levels through ATM/ATR phosphorylation, and the residue has never been mapped, PKA phosphorylates menin at Ser487 and pulls it off chromatin - a kinase input on the drug target itself
528. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P42568 (AF9_HUMAN) canonical FASTA, scanned for [ST]Q motifs 2026-09-19: one match, S395

     · UniProt Consortium 2026
     used by DNA-PKcs phosphorylates AF9 at S395 and releases it from the super elongation complex - the only AF9 regulatory element the fusion keeps, Four regulatory elements on AF9, and the fusion keeps exactly one - the phospho-switch that controls SEC binding, Knock S395A into the MLLT3 half of a KMT2A::MLLT3 model and ask whether it becomes chemotherapy-resistant, Not one phosphorylation site has ever been mapped on a KMT2A fusion protein itself - every site in this beat is an inference across the junction, PRKDC / DNA-PKcs - the kinase that phosphorylates AF9 on the one regulatory residue the fusion keeps
529. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P51530 (DNA2_HUMAN) - DNA replication ATP-dependent helicase/nuclease DNA2, Homo sapiens

     UniProt Consortium · UniProt Knowledgebase 2026
     used by DNA2 (P51530, 1,060 aa) - EXO1's redundant partner in long-range resection, dual mitochondrial and nuclear, and the gene whose DepMap row is measured in 80 screens where every other cascade member has 1,258
530. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch over PubMed, db=pubmed, retmode=json, querytranslation read back for every query

     National Center for Biotechnology Information · NCBI Entrez Programming Utilities 2026
     used by DNA2 (P51530, 1,060 aa) - EXO1's redundant partner in long-range resection, dual mitochondrial and nuclear, and the gene whose DepMap row is measured in 80 screens where every other cascade member has 1,258
531. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ENCODE Portal experiment search, type=Experiment, assay_title=TF ChIP-seq, status=released, target.label=DNA2

     ENCODE Project Consortium · ENCODE Portal 2026
     used by DNA2 (P51530, 1,060 aa) - EXO1's redundant partner in long-range resection, dual mitochondrial and nuclear, and the gene whose DepMap row is measured in 80 screens where every other cascade member has 1,258
532. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P26358 (DNMT1_HUMAN), DNA (cytosine-5)-methyltransferase 1 / CXXC-type zinc finger protein 9

     UniProt Consortium · UniProtKB 2026
     used by DNMT1 - the antagonist: its CXXC domain reads unmethylated CpG in order to STOP methylation, and it is the only heterologous CXXC that runs an MLL-AF9 leukaemia
533. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, queries: (MLL OR KMT2A) AND DNMT3A AND (leukemia OR leukaemia); DNMT3A[tiab] AND (MLL OR KMT2A) AND (rearrang* OR translocation); DNMT3A AND (KMT2A-rearranged OR MLL-rearranged) AND (infant OR pediatric OR paediatric OR childhood)

     National Library of Medicine · PubMed 2026
     used by DNMT3A - an independent adverse factor in adult KMT2A-rearranged AML and conspicuously absent from paediatric disease, which is the whole finding
534. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9Y6K1 (DNM3A_HUMAN), SUBCELLULAR LOCATION comment with evidence codes, sources and the uncoded note, plus TISSUE SPECIFICITY, fetched live 2026-09-21

     UniProt Consortium · Databases 2026
     used by DNMT3A - an independent adverse factor in adult KMT2A-rearranged AML and conspicuously absent from paediatric disease, which is the whole finding
535. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed entry P04053 (TDT_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by DNTT (TdT) - a real gene the graph calls by its enzyme nickname, and TdT-negativity is 100% KMT2A-rearranged in the largest paediatric series
536. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     HGNC record HGNC:2983 (DNTT)

     HUGO Gene Nomenclature Committee · HGNC 2026
     used by DNTT (TdT) - a real gene the graph calls by its enzyme nickname, and TdT-negativity is 100% KMT2A-rearranged in the largest paediatric series
537. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     Q126 cross-partner scan over all fusion entries in this corpus: 3'-partner recurrence and four-axis keyword presence

     Computed by this beat, 2026-09-23 · analyses/q126-the-partners-named-but-never-written/nup98-shared-address-cross-partner-scan.json 2026
     used by Do NUP98's seven fusions resemble each other more than they resemble their partners' other fusions? The test cannot be run for six of them, and the seventh resembles neither side
538. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     fusion-nup98-kmt2a, this corpus (entry read in full this session; its own primary citations are E1/E2/E5 of that entry and were not independently re-fetched here)

     Oncofusion Knowledge Graph, comparative beat · data/graph/nodes/comparative/fusion-nup98-kmt2a.json 2026
     used by Do NUP98's seven fusions resemble each other more than they resemble their partners' other fusions? The test cannot be run for six of them, and the seventh resembles neither side
539. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     mechanism-nup98-fusion-menin-mll1-dependency, this corpus (read in full this session)

     Oncofusion Knowledge Graph, comparative beat · data/graph/nodes/comparative/mechanism-nup98-fusion-menin-mll1-dependency.json 2026
     used by Do NUP98's seven fusions resemble each other more than they resemble their partners' other fusions? The test cannot be run for six of them, and the seventh resembles neither side
540. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q96BY6 (DOC10_HUMAN) - Dedicator of cytokinesis protein 10 (Zizimin-3)

     UniProt Consortium · UniProtKB 2026
     used by DOCK10 (Zizimin-3) - the Cdc42/Rac1 exchange factor fused to KMT2A exon 8 in an undifferentiated pleomorphic sarcoma, and the only one of the seven this graph already held as an H3K79me2-marked gene
541. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     RefSeq NM_014689.3 - Homo sapiens dedicator of cytokinesis 10 (DOCK10), transcript variant 1, mRNA

     NCBI RefSeq · NCBI Nucleotide 2026
     used by DOCK10 (Zizimin-3) - the Cdc42/Rac1 exchange factor fused to KMT2A exon 8 in an undifferentiated pleomorphic sarcoma, and the only one of the seven this graph already held as an H3K79me2-marked gene
542. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paperPMID 40972012
     DOT1L-mediated H3K79me2 directs B-cell repertoire establishment, marginal zone development, and germinal center function.

     Kealy LC, Russ BE, Turner SJ et al · Journal of Immunology 2026
     used by DOCK10 (Zizimin-3) - the Cdc42/Rac1 exchange factor fused to KMT2A exon 8 in an undifferentiated pleomorphic sarcoma, and the only one of the seven this graph already held as an H3K79me2-marked gene, H3K79me2 builds up as B cells mature - so in B-ALL the drug's target is what the rescued compartment runs on, The unnamed Cre is Mb1-Cre, which acts at the pro-B stage: the one resolution in this beat that lands on the right cell
543. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset
     DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan)

     Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
     used by DOCK10 (Zizimin-3) - the Cdc42/Rac1 exchange factor fused to KMT2A exon 8 in an undifferentiated pleomorphic sarcoma, and the only one of the seven this graph already held as an H3K79me2-marked gene
544. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed counted-absence search, menin inhibitors against FANCD2/Fanconi, run 2026-09-20

     · PubMed (NCBI) 2026
     used by Does menin hold FANCD2 in a KMT2A-rearranged cell, and does revumenib let go? Nobody has asked — zero records over every menin-inhibitor name
545. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42063707
     Concurrent therapy-related acute myeloid leukemia and lymph node tuberculosis following treatment for lung squamous cell carcinoma: a case report and literature review.

     Yu Y, Yang D, Chen X, Chen Y · Front Oncol 2026
     used by Does menin hold FANCD2 in a KMT2A-rearranged cell, and does revumenib let go? Nobody has asked — zero records over every menin-inhibitor name
546. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed counted-absence search, menin co-immunoprecipitation against FANCD2/NF-kappaB, run 2026-09-20

     · PubMed (NCBI) 2026
     used by Does menin hold FANCD2 in a KMT2A-rearranged cell, and does revumenib let go? Nobody has asked — zero records over every menin-inhibitor name
547. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41814014
     Immunophenotypic changes following menin inhibition in acute myeloid leukemia.

     Loghavi S, ..., Issa GC · Leukemia 2026
     used by Does the diagnostic marker survive the drug? Surface NG2, CD133, CD10 and CD33 under menin and DOT1L inhibition, with the rearrangement held fixed, Lineage drift - the partial state between B-ALL and AML, and the fact that nobody has published the ORDER in which markers change, NG2 reads the fusion's transcriptional programme, not the DNA lesion - so a menin inhibitor should erase the graph's best diagnostic marker while the rearrangement stays put, What an immunophenotype can and cannot tell you about a genotype - the arithmetic of a surrogate marker, worked on NG2
548. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 1NW3, 3QOW, 3QOX, 3SR4, 6NJ9, 6NQA, 6NOG, 7XCR, 7XCT, 6J99, 6O96, 6NN6, 2MV7, 6JN2 and 7EDP, with per-entity UniProt accessions, methods, resolutions and modelled residue ranges resolved against the RCSB search and data GraphQL APIs

     RCSB Protein Data Bank · RCSB PDB 2026
     used by DOT1L — the target where excellent chemistry met an indirect interaction
549. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, target CHEMBL1795117 (histone-lysine N-methyltransferase, H3 lysine-79 specific): target_search, get_bioactivity and get_mechanism queries

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by DOT1L — the target where excellent chemistry met an indirect interaction
550. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 41925229
     Trans-histone crosstalk establishes distinct H3K79 methylation zones with differential transcriptional functions.

     Park NH, Kim HR, Kim HY, et al. · Nucleic Acids Res 2026
     used by DOT1L methylates H3K79 non-processively - so the three states are one distribution, not three signals, H3K79me1 - the first methyl group, the one H2B ubiquitination is not needed for, and the state HOX loci fall back to as blood cells mature, H3K79me3 - the third methyl group, the one that needs the most enzyme flux and the one human DOT1L makes least of, H4K16ac - the one mark rule 80 actually rescued in this beat: elevated at MLL-AF9 targets in a supplementary figure, and the reason MLL1's own enzyme turns out to be dispensable, KAT8 (MOF) — a genuine KMT2A partner that binds the half of KMT2A the fusion discards, The trans-histone crosstalk H2Bub → DOT1L → H3K79me: modulatory for me1, effectively required for me2/me3
551. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q8TEK3 (DOT1L_HUMAN) - subcellular location with evidence codes, fetched 2026-09-20

     UniProt Consortium · UniProtKB 2026
     used by DOT1L's localization record is a mouse kidney paper wearing a human label - and the one image of it in a leukaemia cell is a supporting figure a title search cannot see
552. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q92785 (REQU_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by DPF2 (BAF45D) - a cBAF-specific reader required in MLL-AF9 cells, with one PubMed record in this disease and no selectivity signal
553. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9C005 (DPY30_HUMAN) and Q9UBL3 (ASH2L_HUMAN) - subcellular location with evidence codes, fetched 2026-09-20

     UniProt Consortium · UniProtKB 2026
     used by DPY30 has a second job at the trans-Golgi network, and the authors who found it could not put ASH2L or RBBP5 there
554. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
     UniProtKB Q9UBX2 (DUX4_HUMAN): sequence length, domain, zinc-finger, DNA-binding, repeat, motif and region features, and PDB cross-references

     UniProt Consortium · UniProtKB 2026
     used by DUX4 (double homeobox 4) domain architecture
555. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB: DUX4 is Q9UBX2 (Double homeobox protein 4, 424 aa, Homo sapiens, verified); the IGH locus has no single UniProtKB accession because it comprises many V, D, J and constant-region genes assembled by somatic recombination

     UniProt Consortium · UniProtKB 2026
     used by DUX4::IGH (IGH-DUX4), t(4;14)/ins(14;4) — D4Z4 repeat insertion into the IGH locus
556. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 41195508
     Menin inhibitor DS-1594b drives differentiation and induces synergistic lethality in combination with venetoclax in acute myeloid leukemia cells with rearranged mixed-lineage leukemia and mutated nucleophosmin-1

     Ciaurro V, Sharlandjieva V, Skwarska A, Chahrour C, Baran N, Zeng Z, et al., Konopleva M · Haematologica 2026
     used by Dynamic BH3 profiling before and after menin inhibition — a one-week assay on primary cells that would choose the BH3-mimetic partner, and has never been run in lymphoid disease, The state under menin inhibition — differentiation in AML, rapid death in ALL, and a GMP-like exit in the cells that switch lineage instead
557. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q96JC9 (EAF1_HUMAN), ELL-associated factor 1

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by EAF1 — the interactor with an actual subnuclear phenotype: an MLL fusion evicts it from Cajal bodies, and the bodies themselves fall apart
558. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q96CJ1 (EAF2_HUMAN), ELL-associated factor 2

     · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
     used by EAF2 (U19) — the EAF paralogue whose ELL-binding domain is a mapped tumour-suppressor module, described entirely in prostate cancer
559. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch record counts for EAF2 and its aliases against KMT2A and MLL, run for this beat

     · NCBI PubMed, queried 2026-09-19 2026
     used by EAF2 (U19) — the EAF paralogue whose ELL-binding domain is a mapped tumour-suppressor module, described entirely in prostate cancer
560. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database, PSICQUIC MITAB 2.7 query on interactor O00255 (MEN1), retrieved 2026-09-20

     · IntAct (EMBL-EBI) 2026
     used by EBF3 — a single large-scale MS hit against menin, in a B-lineage transcription factor family, which is the only reason it is worth recording, HCFC2 — menin's other host cell factor, present in the founding paper by gel filtration and never mentioned again, Menin represses NF-kB transactivation — through a mapped central region, a SIRT1 cofactor, and an RPA2 off-switch, in every tissue except the one this project cares about, Menin-FANCD2: one 2003 co-immunoprecipitation, a curated sufficient region of menin 214-390, and nothing since, Menin's 33 curated partners, enumerated: 10 written, 10 discarded as mutation-directed or single-method, 3 that are not proteins, and what was read and rejected, MYH9 (non-muscle myosin IIA) — the best-validated menin partner nobody in this field talks about, and the only one that maps to menin's N-terminus, NFKB1 (p105/p50) — the only menin-NF-kB subunit with a co-immunoprecipitation, and the arm the MLL leukaemia data points at, NFKB2 (p100/p52) — menin binds the non-canonical subunit too, and activating that arm SUPPRESSES MLL-AF9 leukaemia, RELA (p65) — menin binds it and represses it, and it is required to keep MLL oncoproteins on chromatin, Ten of menin's 33 IntAct partners come from one BRET screen for MUTATION-created interactions — and five are gains caused by a named allele
561. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed counted-absence search for EBF3 against KMT2A/MLL/menin, run 2026-09-20, hit title-checked

     · PubMed (NCBI) 2026
     used by EBF3 — a single large-scale MS hit against menin, in a B-lineage transcription factor family, which is the only reason it is worth recording
562. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q9H4W6 (COE3_HUMAN, Transcription factor COE3)

     · UniProtKB/Swiss-Prot 2026
     used by EBF3 — a single large-scale MS hit against menin, in a B-lineage transcription factor family, which is the only reason it is worth recording
563. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q99848 (EBP2_HUMAN), human EBNA1BP2, 306 aa

     · UniProt Consortium 2026
     used by EBNA1BP2 (EBP2) - the DOT1L spoke with the HIGHEST stoichiometry and the WEAKEST evidence, and the case that shows why abundance is the third dimension
564. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     UniProtKB Q8TEK3 (DOT1L_HUMAN), human DOT1L, 1537 aa

     · UniProt Consortium 2026
     used by EBNA1BP2 (EBP2) - the DOT1L spoke with the HIGHEST stoichiometry and the WEAKEST evidence, and the case that shows why abundance is the third dimension, PARP16 - graded A- in the DOT1L pulldown, the best of the six, and a tail-anchored ER membrane protein that cannot be in the nucleoplasm
565. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed esearch counts for EED-directed inhibitor queries, run and title-checked 2026-09-20

     · NCBI PubMed 2026
     used by EED - the PRC2 subunit this disease actually needs, with drugs in the clinic and zero leukaemia studies, Point the EED inhibitor at the index fusion - a clinical-stage drug, a subunit with no paralogue, and zero leukaemia experiments
566. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB reviewed human record O75530 (EED_HUMAN), resolved with gene_primary and length requested in fields

     · UniProt 2026
     used by EED - the PRC2 subunit this disease actually needs, with drugs in the clinic and zero leukaemia studies
567. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed E-utilities esearch, 2026-09-21. both hits retrieved by efetch and read; the 'nuclear' term in PMID 29363553 belongs to the gene name 'melanoma nuclear protein 18' and not to a compartment

     · NCBI PubMed 2026
     used by EED - the PRC2 subunit this disease actually needs, with drugs in the clinic and zero leukaemia studies
568. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q96KQ7 (EHMT2_HUMAN), Histone-lysine N-methyltransferase EHMT2 - FUNCTION, CATALYTIC ACTIVITY, SUBUNIT, SUBCELLULAR LOCATION, TISSUE SPECIFICITY, DOMAIN and feature table, fetched 2026-09-22

     · UniProt Consortium 2026
     used by EHMT2 / G9a - the H3K9me1/2 writer that sets whether the KMT2A LOCUS ITSELF rearranges, and whose loss delays MLL-AF9 AML in a supplementary figure nobody can find
569. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     UniProtKB - Q9P2K8, E2AK4_HUMAN, eIF-2-alpha kinase GCN2

     · UniProt Knowledgebase 2026
     used by EIF2AK4 / GCN2 (Q9P2K8) - the amino-acid-sensing eIF2alpha kinase: transcript flat, and NOTHING has ever been published about it in any KMT2A-rearranged system under either name
570. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch, db=pubmed, run 2026-09-24; probe-log.jsonl label ledgf/agent-names-any

     · NCBI PubMed 2026
     used by Eleven abstract-level bounded absences re-tested with the instrument that overturned the HDAC6 one - ten survive, one is not checkable this way, none overturned
571. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch, db=pubmed, run 2026-09-24; probe-log.jsonl label enl/yeats-agents

     · NCBI PubMed 2026
     used by Eleven abstract-level bounded absences re-tested with the instrument that overturned the HDAC6 one - ten survive, one is not checkable this way, none overturned
572. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paperPMID 42690622
     Menin inhibition impairs metastatic colonization of Ewing sarcoma.

     · JCI Insight 2026
     used by Eleven abstract-level bounded absences re-tested with the instrument that overturned the HDAC6 one - ten survive, one is not checkable this way, none overturned
573. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     mechanism-kdm3b-depletion-lowers-ctcf-only-at-the-kmt2a-anchor, this graph's own entry written cycle 59, read verbatim 2026-09-22

     · Iqbal Lab oncofusion knowledge graph 2026
     used by Eleven hand-picked intervals are not a null: the CTCF loss on KDM3B depletion scales monotonically with baseline anchor strength (decile medians 0.78 to 0.62), so a strong anchor compared against gene bodies looks unique and compared against its own strength class is the 58th of 1,625, Put three descriptions of the same address on one hg38 axis - the exon-11 CTCF site, the intron-10/exon-11 functional cut point, and the region that breaks more in patients who got a KMT2A fusion - and find out whether they are one element, The KMT2A breakpoint cluster region in base positions: three incompatible definitions reconciled to one GRCh38 table, with the DCAL frame's three bp identities re-derived and the GRCh37 lift measured per interval
574. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     PubMed counted-absence search for DNA damage readouts in primary KMT2A-rearranged blasts, run 2026-09-20

     · PubMed (NCBI) 2026
     used by Eleven immortalised lines passaged for decades: the competing explanation this beat could not close
575. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database (EMBL-EBI), PSICQUIC MITAB 2.7, pubid:35140242, retrieved 2026-09-23

     2026
     used by ELF1 - a direct transcriptional target of MLL fusion proteins and HOXA9 that is ALSO a BioID neighbour of KMT2A, and that is focally deleted in paediatric AML without a KMT2A rearrangement
576. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
     analyses/interactome-tail-size-and-abundance-adjudication/ - orientation.py, pair_null.py, adjudication.json (this repository), pre-registered in PREREGISTRATION.md

     2026
     used by ELF1 - a direct transcriptional target of MLL fusion proteins and HOXA9 that is ALSO a BioID neighbour of KMT2A, and that is focally deleted in paediatric AML without a KMT2A rearrangement, KMT2A-EN1 declined: one spoke-expanded BioID row from an ectopic neural homeobox construct in a kidney cell, and the only screen that ever looked found EN1 unexpressed in AML lines, NFYC / NF-YC - a histone-fold transcription factor whose partner subunit carries an H2B-like ubiquitination, and the most selective of the eleven TF baits that reached KMT2A, NHLH1 / HEN1 / NSCL1 - a TAL1-family bHLH factor found by cross-hybridising with the T-ALL oncogene, whose knockout mouse is normal and whose only imaging call puts it in the GOLGI, Prey recovery in the two BioID depositions tracks SIZE and not abundance - and the interactome tail was being scored on the wrong side of the pair, because every one of the 22 partners is the BAIT and every core protein is the PREY, TBXT / brachyury - the notochord master regulator, and one of only two of 102 transcription-factor baits whose neighbourhood contained DOT1L
577. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     UniProtKB P32519 (ELF1_HUMAN)

     2026
     used by ELF1 - a direct transcriptional target of MLL fusion proteins and HOXA9 that is ALSO a BioID neighbour of KMT2A, and that is focally deleted in paediatric AML without a KMT2A rearrangement
578. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     NCBI PubMed esearch via scripts/pubmed_safe.py, run 2026-09-23; querytranslations recorded: "ELF1"[Title/Abstract] AND "KMT2A"[Title/Abstract] -> 2, and "ELF1"[Title/Abstract] AND "MLL"[Title/Abstract] -> 1

     2026
     used by ELF1 - a direct transcriptional target of MLL fusion proteins and HOXA9 that is ALSO a BioID neighbour of KMT2A, and that is focally deleted in paediatric AML without a KMT2A rearrangement
579. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     Human Protein Atlas v25.1, proteinatlas.json.gz record for ELF1 (ENSG00000120690), retrieved 2026-09-23

     2026
     used by ELF1 - a direct transcriptional target of MLL fusion proteins and HOXA9 that is ALSO a BioID neighbour of KMT2A, and that is focally deleted in paediatric AML without a KMT2A rearrangement
580. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     IntAct molecular interaction database (EMBL-EBI), PSICQUIC tab27 record for id:Q15723 AND id:Q03164

     2026
     used by ELF2 / NERF - an ETS factor that works with RUNX1 on a B-cell kinase promoter, and one of eleven TF baits recovering KMT2A
581. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     UniProtKB Q15723 (ELF2_HUMAN)

     2026
     used by ELF2 / NERF - an ETS factor that works with RUNX1 on a B-cell kinase promoter, and one of eleven TF baits recovering KMT2A
582. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 2DOA (ELL 200-292, NMR) and 2E5N (ELL2 200-292, NMR), with per-entity UniProt accessions, methods and modelled residue ranges resolved against the RCSB APIs; searched for ELL (P55199) complexes with AFF1, AFF4, EAF1, EAF2 and RNA polymerase II subunits

     RCSB Protein Data Bank · RCSB PDB 2026
     used by ELL — rank 7 on selectivity and the deepest common-essential in the top ten, with one NMR fragment to its name
583. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     ChEMBL database version 34, target CHEMBL5725004 (RNA polymerase II elongation factor ELL): target_search, get_bioactivity and get_mechanism queries

     EMBL-EBI ChEMBL · ChEMBL 2026
     used by ELL — rank 7 on selectivity and the deepest common-essential in the top ten, with one NMR fragment to its name, The ELL/EAF activation module (ELL2, EAF1, EAF2) — a general elongation stimulator, not a KMT2A-fusion dependency
584. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB P55199 (ELL_HUMAN) - subcellular location with evidence codes, fetched 2026-09-20

     UniProt Consortium · UniProtKB 2026
     used by ELL is in Cajal bodies - or nuclear speckles, depending which paper you read - and the MLL-ELL fusion empties them
585. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paperPMID 41843233
     Phosphorylation-dependent regulation of Eleven Lysine Rich Leukemia (ELL) uncovered by quantitative global phosphoproteomic analysis.

     Kammarambath SR, et al. · Mol Genet Genomics 2026
     used by ELL S309 is the most frequently observed phosphosite across public datasets - and it is a proline-directed site with no kinase, not an ATM site
586. weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paperPMID 41424126
     Enhanced ELL Phase Separation Is Crucial for Efficient DNA Damage Repair to Restart Transcription and Cell Survival.

     Pal S, et al. · Mol Cell Biol 2026
     used by ELL-EAF1: a phosphorylation-gated interface reported to drive phase separation - and the assays used cannot distinguish that from clustered binding
587. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     RCSB PDB entries 5JW9, 2E5N and 7OKX, verified against the RCSB data API for title, experimental method and resolution

     RCSB Protein Data Bank · RCSB PDB 2026
     used by ELL2 — a SEC subunit that almost no cell needs, next to a paralogue almost every cell needs
588. moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
     PubMed search for ELL2 paralogue specificity in KMT2A/MLL-fusion leukemia, conducted for this graph on 2026-09-18; no such study retrieved

     Oncofusion Knowledge Graph project, using the NCBI PubMed E-utilities · Literature search 2026
     used by ELL2 — a SEC subunit that almost no cell needs, next to a paralogue almost every cell needs
589. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entries O00472 (ELL2_HUMAN) and P55199 (ELL_HUMAN)

     UniProt Consortium · UniProtKB 2026
     used by ELL2 — domain architecture and residue coordinates
590. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     InterPro entry-to-protein matches for UniProt O00472

     InterPro / EMBL-EBI · InterPro 2026
     used by ELL2 — domain architecture and residue coordinates
591. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     SIFTS UniProt residue-level mappings for PDB entries 5JW9, 2E5N (construct ranges read back from the PDBe mappings API)

     PDBe / SIFTS; EMBL-EBI · PDBe SIFTS 2026
     used by ELL2 — domain architecture and residue coordinates
592. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB entry Q9BW60 (ELOV1_HUMAN), Very long chain fatty acid elongase 1

     UniProt Consortium · UniProt 2026
     used by ELOVL1 - the fatty-acid elongase that infant MLL-AF4 ALL retains from its fetal origin, and the lipid enzyme on the opposite side of the divide from ACSL4
593. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9BW60 (ELOV1_HUMAN) SUBCELLULAR LOCATION comment and feature table, fetched 2026-09-21

     UniProt Consortium · UniProtKB 2026
     used by ELOVL1 - the fatty-acid elongase that infant MLL-AF4 ALL retains from its fetal origin, and the lipid enzyme on the opposite side of the divide from ACSL4
594. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB O95163 (ELP1_HUMAN) - reviewed human record, length 1332 aa

     UniProt Consortium · UniProtKB 2026
     used by ELP1 (IKBKAP/IKAP) - the Elongator scaffold that was first named for the IKK complex, is a paediatric-cancer predisposition gene, and has an RNA Pol II contact in IntAct
595. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch over PubMed and GEO DataSets, run 2026-09-22 by Q98 with querytranslation asserted non-empty on every call

     Oncofusion knowledge-graph project (Q98-elongator-smyd3) · NCBI E-utilities 2026
     used by ELP1 (IKBKAP/IKAP) - the Elongator scaffold that was first named for the IKK complex, is a paediatric-cancer predisposition gene, and has an RNA Pol II contact in IntAct
596. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     IntAct molecular interaction database, REST endpoint findInteractions (not PSICQUIC)

     IntAct/EMBL-EBI · IntAct 2026
     used by ELP1 (IKBKAP/IKAP) - the Elongator scaffold that was first named for the IKK complex, is a paediatric-cancer predisposition gene, and has an RNA Pol II contact in IntAct, ELP2 (STATIP1/StIP1) - the double beta-propeller of the Elongator core, whose informal name collides with a different human gene and whose STAT3 function is a by-similarity annotation, ELP3 - the catalytic subunit, the only Elongator protein with a real acetyltransferase domain, and the one whose loss in mouse blood causes marrow failure and then p53-mutant leukaemia, ELP5 (DERP6) - the bridge that joins ELP3 to ELP4, the most essential Elongator subunit in DepMap, and a gene whose cytoplasmic call rests on an overexpression experiment, ELP6 (TMEM103/C3orf75) - an Elongator subunit named after a membrane protein it is not, whose entire localisation record comes from a Drosophila paper its own symbol cannot retrieve
597. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q6IA86 (ELP2_HUMAN) - reviewed human record, length 826 aa

     UniProt Consortium · UniProtKB 2026
     used by ELP2 (STATIP1/StIP1) - the double beta-propeller of the Elongator core, whose informal name collides with a different human gene and whose STAT3 function is a by-similarity annotation
598. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch over PubMed with esummary title-checking, run 2026-09-22 by Q98 with querytranslation asserted non-empty on every call

     Oncofusion knowledge-graph project (Q98-elongator-smyd3) · NCBI E-utilities 2026
     used by ELP2 (STATIP1/StIP1) - the double beta-propeller of the Elongator core, whose informal name collides with a different human gene and whose STAT3 function is a by-similarity annotation
599. strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
     NCBI E-utilities esearch over PubMed, run 2026-09-22 by Q98 with querytranslation asserted non-empty on every call

     Oncofusion knowledge-graph project (Q98-elongator-smyd3) · NCBI E-utilities 2026
     used by ELP2 (STATIP1/StIP1) - the double beta-propeller of the Elongator core, whose informal name collides with a different human gene and whose STAT3 function is a by-similarity annotation
600. established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
     UniProtKB Q9H9T3 (ELP3_HUMAN) - reviewed human record, length 547 aa

     UniProt Consortium · UniProtKB 2026
     used by ELP3 - the catalytic subunit, the only Elongator protein with a real acetyltransferase domain, and the one whose loss in mouse blood causes marrow failure and then p53-mutant leukaemia