Oncofusion

start / graph / absence-no-h3k9-writer-or-reader-localised-at-the-kmt2a-locus

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q110-h3k9me3-writer-layer

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

also called no writer at the KMT2A locus · H3K9 writer occupancy absence · SUV39H1 has no ENCODE ChIP-seq

This graph has argued for several cycles about a silencing tag on the KMT2A gene. Nobody has ever measured where the enzymes that write that tag, or the protein that reads it, actually sit on that gene - in any cell, in any published paper. For the main writer there is not even a public dataset that could be checked.

The same thing, for a scientist

**CORRECTED 2026-09-22 (cycle 66) BY Q114, AND THE CORRECTION IS TO THIS ENTRY'S HEADLINE, NOT A FOOTNOTE.** This entry classified SUV39H1 as `never-measured` and said *for SUV39H1 no archive can be asked either*. **Both halves are now wrong, in two steps.** Q113 (cycle 65) found two GEO deposits holding six human SUV39H1 occupancy samples in two cell types. Q114 (this cycle) **read them at this locus**. SUV39H1's absence class is therefore `measured-not-analysed` as of cycle 65 and **`measured`** as of cycle 66. **What the reading found, which does not change this entry's conclusion but changes its standing entirely:** in SU-DIPG13 and SU-DIPG36 (GSE225028, CUT&RUN, hg38 bigWig) SUV39H1 is **NOT OCCUPIED** at any of the seven KMT2A intervals - the 147 bp exon-11 anchor reads an **explicit zero** in SU-DIPG13 and 1.666 (74.2nd percentile of 59,821 class-matched CTCF anchors) in SU-DIPG36, and no interval in either line exceeds 1.21x the genome-wide random-tile mean. So the *conclusion* that no H3K9 writer has been placed at this locus survives for the published literature, and it is now backed by a direct measurement rather than by the absence of one. **The ENCODE half of the census is verified and stands**: three independent routes on 2026-09-22 - `/targets/SUV39H1-human/` HTTP 404, `type=Target&searchTerm=SUV39H1` total 0, `type=Experiment&target.label=SUV39H1` total 0 - with `/targets/BARD1-human/` returning 200 and `/targets/EXO1-human/` 404 as controls through the identical URL template. **The word that was wrong was `never`, and the lesson is the registry's own: state the archive in the claim.** See `ana-q114-every-route-to-human-suv39h1-occupancy-data-leads-to-the-same-two-deposits` and `ana-q114-suv39h1-is-not-at-the-kmt2a-locus-in-either-patient-derived-glioma-line`. --- **The question, asked as narrowly as it can be asked:** has SUV39H1, SETDB1 or HP1-alpha/CBX5 ever been placed at the *KMT2A* locus, in any cell type? **No.** **Twenty PubMed queries, five alias spellings, zero on-topic records** [E1]. Every conjunction with a locus term returns 0 - `SUV39H1[tiab] AND 11q23[tiab]`, `SETDB1[tiab] AND 11q23[tiab]`, `SUV39H1[tiab] AND KMT2A[tiab] AND locus[tiab]`, `SUV39H1[tiab] AND "fragile site"[tiab]`, `CBX5[tiab] AND KMT2A[tiab]` - and every conjunction that returns a small non-zero count title-checks to nothing: the three `SUV39H1 AND KMT2A` records are a DOT1L/SIRT1 paper already in this graph, a **mangrove rivulus fish** gene-family survey and an adipose-stem-cell spheroid paper; the two `SETDB1 AND KMT2A` records are the same fish paper and an NMDA-receptor chromatin paper; the single `SUV39H1 AND breakpoint` record is about the late-S cell-cycle breakpoint; the single `SETDB1 AND breakpoint` record is about the **BCR gene of BCR::ABL1**, not the *KMT2A* breakpoint cluster region. **The alias widening was run in both directions and it changed nothing here** [E1]. `KMT1A[tiab]` returns 22 records on its own and `KMT1A[tiab] AND MLL[tiab]` returns **0**; `ESET` and `KMT1E` against MLL both return **0**; HP1-alpha was run in five spellings (`CBX5`, `HP1alpha`, `HP1a`, `"HP1 alpha"`, `heterochromatin protein 1`) plus the bare `HP1`, and the only on-topic record in the whole set is a **2001 Drosophila** paper about dAF10 and fly HP1. **The bare `HP1` operand is contaminated**: `HP1[tiab] AND MLL[tiab]` returns 3 and one of them is *Klebsiella oxytoca* strain HP1 from a hot spring. **Rule 67 is satisfied - the query shape demonstrably returns ones.** `SUV39H1[tiab] AND menin[tiab]` = 7, `SUV39H1[tiab] AND CTCF[tiab]` = 6, `SETDB1[tiab] AND CTCF[tiab]` = 11, `SETDB1[tiab] AND leukemia[tiab]` = 30, `HP1alpha[tiab] AND leukemia[tiab]` = 18. Every zero above has a non-empty `querytranslation` and `warninglist` *"No items found."* **The three absences are NOT the same class, and that is the operationally useful finding** [E2]: | protein | literature | ENCODE ChIP-seq | absence class | what closes it | |---|---|---|---|---| | **SUV39H1** | 0 | **0 experiments, ANY cell type** | `never-measured` | new wet work, or a GEO sweep this beat did not run | | **SETDB1** | 0 | 3 experiments, all **K562** | `measured-not-analysed` | compute - done in `ana-q110-h3k9-writer-and-reader-occupancy-at-the-kmt2a-locus` | | **CBX5** | 0 | 4, two in blood: **K562 and GM12878** | `measured-not-analysed` | compute - same analysis | So the writer this graph's dispute is actually about is the one member of the machinery no archive can answer for, while the other two were answerable all along and nobody had asked. **What this does and does not license.** It does not say H3K9 methylation is irrelevant at this locus: Q108 measured the MARK there, in RPE, and a blood-cell H3K9me1 track replicates the BCR-scale depletion. It says that **every published statement assigning that mark at that element to a particular enzyme is an inference from enzyme perturbation, never from enzyme occupancy** - and `gene-asb7`, `gene-suv39h1` and `gene-cbx5` between them now record three separate reasons why a mark level can move without the enzyme's activity changing.

Facts and figures
scriptno script - this is a query and archive census, with every count and querytranslation recorded inline; the companion computation is analyses/q110-h3k9-writer-and-reader-at-the-kmt2a-locus/run.py
statusrun
run date2026-09-22
run byQ110-h3k9me3-writer-layer
absence classmeasured-not-reported
what is absentAny measurement of SUV39H1, SETDB1 or CBX5/HP1-alpha OCCUPANCY at the KMT2A gene, its 7,466 bp breakpoint cluster region, or its exon-11 CTCF/cohesin element, in any cell type, in the published literature.
what is NOT absentMeasurements of the MARK at this locus exist - GSE210480's H3K9me1/me2/me3 in RPE (Q108) and ENCODE's K562 H3K9me1 (the cycle-59 king analysis). This absence is about the ENZYMES and the READER, which is a different and previously unasked question.
why this absence is load bearingmechanism-h3k9me1-2-ctcf-balance-gates-kmt2a-rearrangement models the H3K9 state at this locus as a KDM3B-versus-G9a balance and prescribes a G9a/GLP inhibitor prophylactically. mark-h3k9me3 argues the opposite sign from SETDB1 and SUV39H1 biology at fusion-target promoters. Both arguments assign the mark to enzymes, and NEITHER enzyme has been observed at the locus in question. Three separate mechanisms recorded in this graph let the mark move without the enzyme's activity changing: CUL5-ASB7 degrading SUV39H1 (gene-asb7), lamin A protecting it (gene-suv39h1 E11), and G2E3/H3K14ub licensing where it goes (gene-suv39h1 E12).
localizationchromatin at chr11q23.3 - the KMT2A gene (GRCh38 chr11:118,436,492-118,526,832), its 7,466 bp DCAL breakpoint cluster region and its 1,048 bp exon-11 CTCF/RAD21/SMC3 element. The absence is an absence AT that address.
subnuclear localizationnot measured for any of the three proteins at this locus in any cell type. Each has a well-documented subnuclear compartment elsewhere - SUV39H1 pericentromeric, laminar and nucleolar; SETDB1 non-pericentromeric and nucleolus-excluded; CBX5 centromeric and pericentromeric heterochromatin - and those compartments DISAGREE with each other, which is why an occupancy measurement at a transcribed euchromatic loop anchor would be informative rather than confirmatory.
cell typethe absence spans every cell type: no report in any human or murine cell, and the archive route reaches only K562, GM12878, H1, HepG2, HEK293, HCT116 and A549 - of which two are blood cells and none is KMT2A-rearranged. Nothing in a CD34+ progenitor and nothing in infant material.
cell stateunmeasured in every state, and state matters here more than usual: SUV39H1 and SETDB1 are both recruited to chromatin transiently on DNA damage and released when ATM phosphorylates KAP1, so a resting-state occupancy null at a break-prone element is weak evidence of absence and a damage-state measurement has never been attempted at this locus.
epigenetic readoutthe absent measurement is writer and reader OCCUPANCY of H3K9me3-associated proteins; the mark itself (H3K9me1, me2, me3) has been measured here and is recorded in mark-h3k9me1, mark-h3k9me2, mark-h3k9me3 and the Q108 analyses.
shown for fusionNO - nobody has looked, in any fusion, for occupancy of any H3K9 writer or reader at the KMT2A locus. The MLL-fusion results that exist for all three proteins are at FUSION-TARGET genes (Hoxa9, Meis1) or are expression and immortalisation phenotypes, never at the KMT2A gene itself.
absence class noteMIGRATED 2026-09-23. This field previously held the whole verdict as free text: 'literature-absence (bounded by titles and abstracts); the ARCHIVE half is now measured - see corrected_2026_09_22'. That string was out of the controlled vocabulary, so a checker could not read it - but it said more than any token can, and it has been kept here rather than deleted. The original free text read 'literature-absence (bounded by titles and abstracts); the ARCHIVE half is now measured - see corrected_2026_09_22'. Both halves are real and they have different classes: the LITERATURE half is a bounded literature absence, and the ARCHIVE half was measured by the cycle-66 beat and came back negative. The single token that fits the residue is measured-not-reported - public samples exist, the analysis has now been run here, and no published paper reports it.
absence class per protein
{
  "SUV39H1": "CORRECTED 2026-09-22. Previously 'never-measured'. Now MEASURED: ENCODE holds zero experiments (three routes, controlled, re-verified 2026-09-22), but GEO holds GSE225028 (CUT&RUN, SU-DIPG13 and SU-DIPG36) and GSE139196 (ChIP-seq, MCF-7 +/- estradiol), six samples in two human cell types, and Q114 read all of them at this locus. SUV39H1 is NOT OCCUPIED at any of the seven KMT2A intervals in either DIPG line.",
  "SETDB1": "measured-not-analysed - three ENCODE K562 experiments exist and were never read at this locus until this beat",
  "CBX5": "measured-not-analysed - four ENCODE experiments exist, two in blood cells, never read at this locus until this beat"
}
queries run 2026 09 22
{
  "instrument": "NCBI E-utilities esearch over PubMed, every querytranslation read back and every zero checked for a non-empty translation plus warninglist 'No items found.'",
  "query_form": "one [tiab] tag per term, never the grouped (A OR B)[tiab] form (rule 94); no unquoted cytogenetic parentheses anywhere (rule 95)",
  "locus_conjunctions_all_zero": {
    "SUV39H1[tiab] AND 11q23[tiab]": 0,
    "SETDB1[tiab] AND 11q23[tiab]": 0,
    "SUV39H1[tiab] AND KMT2A[tiab] AND locus[tiab]": 0,
    "SUV39H1[tiab] AND \"fragile site\"[tiab]": 0,
    "CBX5[tiab] AND KMT2A[tiab]": 0,
    "SUV39H1[tiab] AND \"MLL-AF4\"[tiab]": 0,
    "SETDB1[tiab] AND \"MLL-AF4\"[tiab]": 0,
    "CBX5[tiab] AND \"MLL-AF9\"[tiab]": 0,
    "SUV39H1[tiab] AND infant[tiab]": 0,
    "SUV39H1[tiab] AND etoposide[tiab]": 0,
    "SETDB1[tiab] AND \"acute lymphoblastic leukemia\"[tiab]": 0,
    "CBX5[tiab] AND \"hematopoietic stem\"[tiab]": 0
  },
  "small_nonzero_counts_all_title_checked_to_zero_on_topic": {
    "SUV39H1[tiab] AND KMT2A[tiab]": {
      "count": 3,
      "records": {
        "25822366": "DOT1L inhibits SIRT1-mediated epigenetic silencing... - already in this graph, H3K9me2 at fusion-bound TSSs, NOT at the KMT2A locus",
        "30458291": "The Kdm/Kmt gene families in the self-fertilizing mangrove rivulus fish, Kryptolebias marmoratus",
        "38173411": "Histone Trimethylations and HDAC5 Regulate Spheroid Subpopulation... of Human Adipose-Derived Stem Cells"
      }
    },
    "SETDB1[tiab] AND KMT2A[tiab]": {
      "count": 2,
      "records": {
        "25467983": "Conserved higher-order chromatin regulates NMDA receptor gene expression and cognition",
        "30458291": "the same mangrove rivulus fish survey"
      }
    },
    "SUV39H1[tiab] AND breakpoint[tiab]": {
      "count": 1,
      "records": {
        "30312172": "H3K9me3 and H4K20me3 represent the epigenetic landscape for 53BP1 binding to DNA lesions - 'breakpoint' here is the late-S CELL-CYCLE breakpoint, not a genomic breakpoint"
      }
    },
    "SETDB1[tiab] AND breakpoint[tiab]": {
      "count": 1,
      "records": {
        "42055326": "HIRA-SETDB1-H3K9me3 axis regulates chromatin architecture in leukemia cells - 'Breakpoint Cluster Region protein' there is the BCR GENE of BCR::ABL1, not the KMT2A breakpoint cluster region"
      }
    },
    "SUV39H1[tiab] AND translocation[tiab]": {
      "count": 1,
      "records": {
        "38976080": "Nuclear translocation of ISG15 regulated by PPP2R2B inhibits cisplatin resistance of bladder cancer - 'translocation' is protein translocation"
      }
    },
    "CBX5[tiab] AND MLL[tiab]": {
      "count": 1,
      "records": {
        "19200802": "Hierarchical maintenance of MLL myeloid leukemia stem cells... - a real and important MLL result (see gene-cbx5) but an expression and immortalisation result, NOT an occupancy measurement"
      }
    },
    "CBX5[tiab] AND CTCF[tiab]": {
      "count": 1,
      "records": {
        "42204197": "Prognostic significance of IDH1 promoter methylation and associated genome-wide alterations in breast cancer"
      }
    },
    "HP1[tiab] AND MLL[tiab]": {
      "count": 3,
      "records": {
        "11266362": "The Drosophila homolog of the human AF10 is an HP1-interacting suppressor of position effect variegation - FLY HP1, fly dAF10",
        "15636750": "Isolation and characterization of a high H2-producing strain Klebsiella oxytoca HP1 from a hot spring",
        "17967446": "Antagonistic functions of SET-2/SET1 and HPL/HP1 proteins in C. elegans development"
      }
    },
    "HP1[tiab] AND KMT2A[tiab]": {
      "count": 1,
      "records": {
        "27895715": "Protein kinase Msk1 physically and functionally interacts with the KMT2A/MLL1 methyltransferase complex - HP1 appears only in the authors' proposed model of chromatin condensation, with no HP1 experiment"
      }
    }
  },
  "alias_widening_run_in_both_directions": {
    "SUV39H1": {
      "KMT1A[tiab] alone": 22,
      "KMT1A[tiab] AND MLL[tiab]": 0,
      "verdict": "the alias reaches a real literature of its own and NOTHING on this conjunction - so on this gene the widening resolved to no, and the answer was not free to assume"
    },
    "SETDB1": {
      "ESET[tiab] AND MLL[tiab]": 0,
      "KMT1E[tiab] AND MLL[tiab]": 0,
      "verdict": "both aliases sterile on this conjunction - the OPPOSITE of the registry's EHMT2/G9a row, where the retired alias is the better operand. The alias question still has no default direction."
    },
    "CBX5": {
      "CBX5[tiab] AND leukemia[tiab]": 4,
      "HP1alpha[tiab] AND leukemia[tiab]": 18,
      "\"HP1 alpha\"[tiab] AND leukemia[tiab]": 3,
      "verdict": "HERE THE WIDENING MATTERED, four-fold, for the DISEASE conjunction - the common name reaches 18 records where the approved symbol reaches 4 - and it still returned nothing at the locus. Both were run and both are recorded."
    }
  },
  "rule_67_positive_controls_through_the_identical_shape": {
    "SUV39H1[tiab] AND menin[tiab]": 7,
    "SUV39H1[tiab] AND CTCF[tiab]": 6,
    "SUV39H1[tiab] AND MLL[tiab]": 8,
    "SETDB1[tiab] AND CTCF[tiab]": 11,
    "SETDB1[tiab] AND leukemia[tiab]": 30,
    "SETDB1[tiab] AND TRIM28[tiab]": 54,
    "HP1alpha[tiab] AND leukemia[tiab]": 18,
    "verdict": "the shape returns ones freely, so the zeros above are not an instrument failure"
  },
  "bound_class": "TITLES AND ABSTRACTS ONLY (rule 80). A SUV39H1, SETDB1 or CBX5 ChIP track crossing chr11:118.43-118.53 Mb in a supporting figure of a paper about something else would be invisible to every query here, and the measurements this graph most wants are usually supporting panels. This bound is therefore a bound on what has been REPORTED IN AN ABSTRACT, and nothing stronger."
}
archive census 2026 09 22
{
  "instrument": "ENCODE Portal experiment search, type=Experiment, assay_title=TF ChIP-seq, status=released, one query per target label",
  "SUV39H1": {
    "n": 0,
    "biosamples": []
  },
  "SUV39H2": {
    "n": 0,
    "biosamples": []
  },
  "SETDB1": {
    "n": 5,
    "biosamples": [
      "K562",
      "K562",
      "K562",
      "HepG2",
      "HEK293"
    ]
  },
  "CBX5": {
    "n": 4,
    "biosamples": [
      "K562",
      "GM12878",
      "H1",
      "HepG2"
    ]
  },
  "CBX1": {
    "n": 2,
    "biosamples": [
      "K562",
      "HepG2"
    ]
  },
  "CBX3": {
    "n": 4,
    "biosamples": [
      "K562",
      "K562",
      "GM12878",
      "HCT116"
    ]
  },
  "EHMT2": {
    "n": 3,
    "biosamples": [
      "K562",
      "A549",
      "HepG2"
    ]
  },
  "TRIM28": {
    "n": 5,
    "biosamples": [
      "K562",
      "K562",
      "K562",
      "HEK293",
      "HEK293"
    ]
  },
  "the_asymmetry_that_matters": "every member of the H3K9 machinery this graph names is represented in ENCODE EXCEPT the SUV39H paralogues, and SUV39H1 is the one the dispute is about. It is also the one gene_asb7's circuit turns on.",
  "what_this_bound_does_not_cover": "individual-lab depositions in GEO. Not swept. One GEO query could move SUV39H1's archive half from never-measured to measured-not-analysed, and that is the cheapest next step in this beat.",
  "assay_title_caveat": "'TF ChIP-seq' is the ENCODE assay title under which these chromatin proteins are filed; an experiment filed under a different assay title would be missed by this exact query form."
}
what it would take to close each half
{
  "the_SETDB1_and_CBX5_half": "CLOSED BY COMPUTE THIS CYCLE, and the answer is no. ana-q110-h3k9-writer-and-reader-occupancy-at-the-kmt2a-locus read ten conservative-IDR peak sets and fourteen signal tracks in K562 and GM12878: ZERO called peaks for SETDB1, CBX5, CBX1, CBX3 or TRIM28 anywhere in the 90 kb KMT2A gene, in the BCR, at the exon-11 element, or within 50 kb of the gene - with the ZNF274 GENE-BODY positive control passing. NOTE on that control: the first run used a window inherited from a sibling analysis and labelled ZNF274 at chr19:57,996,000-58,000,000, which is 187 kb from the real gene in GRCh38 and 698 kb in GRCh37; resolving ZNF274 live (ENSG00000171606, GRCh38 chr19:58,183,029-58,215,964) moved the control from FAIL to PASS.",
  "the_SUV39H1_half": "a GEO sweep first (free), then new ChIP - and the natural experiment is the one this graph already holds as exp-h3k9me3-and-ctcf-at-the-kmt2a-anchor-in-a-cd34-progenitor, extended from the marks to the enzymes",
  "the_lineage_half_that_stays_open_regardless": "K562 is CML-derived erythroleukaemia and GM12878 is an EBV-immortalised lymphoblastoid line. Neither is a CD34+ progenitor and NEITHER IS KMT2A-REARRANGED. No archive holds H3K9 writer or reader occupancy in a KMT2A-rearranged cell, and that half needs wet work.",
  "and_one_thing_that_opened": "G9a/EHMT2 and TRIM28 DO occupy this locus in K562 - four EHMT2 peaks clustered at the KMT2A promoter, one 20 kb upstream shared with TRIM28, and one TRIM28 peak at the promoter - and none of them is at the breakpoint cluster region or the exon-11 element, the nearest being 44.6 kb away. Those are the first and only locus-resolved H3K9-machinery occupancy facts at the KMT2A locus in any cell type, and G9a is the enzyme mechanism-h3k9me1-2-ctcf-balance-gates-kmt2a-rearrangement names."
}
corrected 2026 09 22
{
  "by": "Q114-suv39h1-at-the-locus",
  "what_was_wrong": "the word 'never' in 'SUV39H1: never-measured' and the clause 'for SUV39H1 no archive can be asked either'",
  "what_is_right": "ENCODE holds zero SUV39H1 experiments under any label (three routes, positive and negative controls through the identical URL template, 2026-09-22); GEO holds two deposits with six SUV39H1 samples in two human cell types; both have now been read at this locus",
  "what_survives_unchanged": "the 20-query literature bound, its alias widening, its rule-67 positive controls, and the finding that every published statement assigning H3K9 methylation at a locus to a particular enzyme is an inference from enzyme PERTURBATION rather than from enzyme OCCUPANCY",
  "measurement_that_replaces_the_absence": "ana-q114-suv39h1-is-not-at-the-kmt2a-locus-in-either-patient-derived-glioma-line",
  "archive_census_that_bounds_it": "ana-q114-every-route-to-human-suv39h1-occupancy-data-leads-to-the-same-two-deposits",
  "still_absent": "any SUV39H1, SETDB1 or CBX5 occupancy measurement in a haematopoietic cell, in a KMT2A-rearranged cell, or in a damage state - in any archive"
}
How this could bite us back
  • This absence is the reason no H3K9 intervention in this graph can be signed. mechanism-h3k9me1-2-ctcf-balance-gates-kmt2a-rearrangement prescribes a G9a/GLP inhibitor prophylactically on the strength of an enzyme-balance model of the mark at this locus, and neither enzyme in that model, nor either H3K9me3 writer, nor the reader, has ever been observed at the locus. A prophylactic drug proposal for children should not rest on an occupancy assumption nobody has tested.
  • Do not let this absence be read as permission either. 'No writer has been localised there' is not 'the writers are not there' - it is 'nobody looked', and for SUV39H1 'nobody could have looked from public data'. An argument that lowering H3K9 methylation is therefore safe at this locus would be building on the same missing measurement from the other side.
  • The cheapest closing step is a compute step and it substitutes the wrong cell. Reading K562 and GM12878 answers the blood-cell question and not the disease question, and a null in a non-KMT2A-rearranged line licenses nothing about a KMT2A-rearranged blast - the identical error risk-the-h3k9me3-exoneration-of-g9a-inhibition-is-epithelial-only exists to block for RPE.
Still open
  1. Does GEO hold a SUV39H1 ChIP-seq experiment in any human cell? One esearch against GEO DataSets, or a ChIP-Atlas lookup, moves this from never-measured to measured-not-analysed and is the cheapest unrun step in the whole beat. This beat did not run it. ANSWERED IN CYCLE 65 BY Q113 AND THEN READ IN CYCLE 66 BY Q114. GEO holds two deposits: GSE225028 (SUV39H1 CUT&RUN in SU-DIPG13 and SU-DIPG36, hg38 bigWig) and GSE139196 (four SUV39H1 ChIP-seq in MCF-7 +/- estradiol, hg38 bedGraph, indexed by ChIP-Atlas as SRX7030945-48). ChIP-Atlas holds nothing beyond the latter - its entire hg38 SUV39H1 target-gene table is those same four samples, 26 genes, and KMT2A is not one of them. Q114 read both deposits at this locus: SUV39H1 is NOT OCCUPIED in either DIPG line.
  2. Would a full-text search reach an H3K9 writer track over KMT2A in a supporting figure? Europe PMC fullTextXML reaches inside papers where esearch cannot, and rule 80 says the measurements this graph wants are usually supporting panels. No candidate paper was identified to open, which is itself the difficulty.
  3. Has anyone measured H3K9 writer or reader occupancy at the KMT2A locus in a KMT2A-REARRANGED cell? That question is unanswerable from any archive, because no archive holds a KMT2A-rearranged line for any of these three targets, and it is a different question from the progenitor one.
  4. Is a DAMAGE-STATE occupancy measurement the right experiment rather than a resting one? Both writers are recruited to breaks transiently by CTCF and released on ATM-mediated KAP1 phosphorylation, so the resting-state question this beat answered may be the less informative of the two.
  5. Can the ASB7 CDK1 phosphosite be obtained at all? The paper is paywalled and route 5 returns a Cloudflare challenge. Until it is read, this graph cannot write gene-cdk1 with anything specific to say, and a generic cell-cycle-kinase entry would be worse than none.
  6. NEW, FROM THE CORRECTION (cycle 66): if SUV39H1 is measurable and absent at this locus in the only two human systems that exist, is the H3K9me3 this graph argues about here deposited by SETDB1, by EHMT2/G9a, or by nothing - and which of those three has ever been measured at the locus in a cell of the right lineage? SETDB1 and CBX5 have been read in K562 and GM12878 only; EHMT2 has never been read here at all.
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    Twenty PubMed queries across five alias spellings return zero on-topic records for SUV39H1, SETDB1 or CBX5/HP1-alpha occupancy at the KMT2A locus. Every locus conjunction returns a hard zero with a non-empty querytranslation and warninglist 'No items found.' - including SUV39H1 AND 11q23, SETDB1 AND 11q23, SUV39H1 AND KMT2A AND locus, SUV39H1 AND 'fragile site' and CBX5 AND KMT2A - and every small non-zero count title-checks to nothing on topic, the residue being a mangrove rivulus fish gene-family survey, an NMDA-receptor chromatin paper, an adipose-stem-cell spheroid paper, a cell-cycle-breakpoint DNA-lesion paper, a BCR::ABL1 chromatin-architecture paper whose 'Breakpoint Cluster Region' is a different gene, an ISG15 protein-translocation paper, a breast-cancer methylation paper, a C. elegans development paper and a hydrogen-producing Klebsiella oxytoca strain named HP1. The alias widening was run in both directions for all three genes: KMT1A alone returns 22 records and KMT1A AND MLL returns 0; ESET AND MLL and KMT1E AND MLL both return 0; and for CBX5 the common name reaches four times as much disease literature as the approved symbol (HP1alpha AND leukemia 18 versus CBX5 AND leukemia 4) and still nothing at the locus. Positive controls through the identical query shape return 6 to 54 records.

    What this evidence cannot show

    RULE 80 APPLIES AND IS THE PRINCIPAL LIMIT: this bounds titles and abstracts, not papers. An occupancy track over chr11:118.43-118.53 Mb sitting in a supporting figure of a paper about something else is invisible to every query here, and this project has already paid for that exact failure once - a SET-NUP214 H3K79me2 ChIP across five HOXA promoters that was in Figure 6A and in no abstract. No full-text search was run, and the two or three papers most likely to contain such a panel as a supporting figure were NOT opened, because no candidate paper exists: there is no paper about H3K9 writers at this locus to open.

    Evidence that argues against this (2)
    • The counter-example this beat found by accident is the shape of what could still be hiding. PMID 19881540 places SUV39H1 by ChIP at the p15INK4B and E-cadherin promoters in AML cells, and that fact is reachable only because the promoters are named in the abstract. A paper that ran a SUV39H1 ChIP-seq and mentioned only its headline locus would contribute a genome-wide track covering KMT2A that no query in this entry can see.
    • A negative literature search is weaker than a negative archive query, and this graph has recorded that lesson before. The archive half of this absence [E2] is the stronger half, and it was run - but only against one archive.
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    ENCODE holds zero released ChIP-seq experiments for SUV39H1 and zero for SUV39H2 in any biosample, while every other member of the H3K9 machinery this graph names is represented: SETDB1 5 experiments (K562 x3, HepG2, HEK293), CBX5 4 (K562, GM12878, H1, HepG2), CBX1 2 (K562, HepG2), CBX3 4 (K562 x2, GM12878, HCT116), EHMT2 3 (K562, A549, HepG2) and TRIM28 5 (K562 x3, HEK293 x2). So the three absences are of different classes: SUV39H1 is never-measured because no archive route exists, while SETDB1 and CBX5 were measured-not-analysed, with CBX5 uniquely represented in TWO blood cells including the lymphoblastoid line GM12878.

    What this evidence cannot show

    This bounds ENCODE and nothing else. GEO individual-lab depositions were NOT swept for SUV39H1 ChIP-seq, so a stricter reading of SUV39H1's archive half is measured-not-searched rather than never-measured, and one GEO query could change it - that is named here as the cheapest unrun step. 'TF ChIP-seq' is the assay title these proteins are filed under; a different assay title would be missed. No claim is made about what the SETDB1 or CBX5 tracks show - that is the companion analysis.

    Evidence that argues against this (2)
    • An absence of linkage or of listing in one metadata graph is not an absence of data - this project's own registry records a case where a GEO SuperSeries reported no raw data three different ways and had 79 SRA experiments. ENCODE returning zero for SUV39H1 means ENCODE has not run it, and says nothing about ChIP-Atlas, GEO or a paper's own deposition.
    • Even the positive half is a substitution. K562 is CML-derived erythroleukaemia, GM12878 is EBV-immortalised lymphoblastoid, and neither is a CD34+ progenitor or KMT2A-rearranged - so closing the SETDB1 and CBX5 halves by compute answers 'is the writer or reader at this locus in a blood cell', not 'is it there in the disease'.
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The founding paper for the CUL5-ASB7 circuit that makes this absence load-bearing is itself unreadable beyond its abstract, which is why the CDK1 phosphosite on ASB7 cannot be recorded in this graph. PMID 40440427 has no PMCID, Europe PMC reports isOpenAccess N with a 1,034-character abstract, and resolving its DOI with a browser User-Agent returns HTTP 403 and a 5,722-byte body that strips to 60 characters of Cloudflare challenge text ('Just a moment... Enable JavaScript and cookies to continue'). All five routes this project recognises therefore fail.

    ASB7 is a negative regulator of H3K9me3 homeostasis. · Zhou J, Chen Y, et al. · Science 2025
    PMID 40440427DOI
    What this evidence cannot show

    The author string here is reproduced from this graph's existing gene-asb7 entry rather than re-fetched from PubMed, and is marked as such; the PMID, DOI and title were confirmed against the live Europe PMC core record in this session. The readability test was run at the coordinator's own network; an institutional proxy might succeed where this failed.

    Evidence that argues against this (1)
    • Unreadability is a property of this session's access, not of the paper. Route 5 failing with a Cloudflare challenge is exactly the failure mode that a different network, a library proxy or an author request would defeat, and the phosphosite is presumably named in the paper's figures. This is recorded as an unreadable-literature item, not as an absence of the fact.
Connected to