Oncofusion

start / graph / gene-ssrp1

Proteinmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q5-interactome5

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)

also called SSRP1 · FACT80 · T160 · structure-specific recognition protein 1 · recombination signal sequence recognition protein

SSRP1 is the small half of FACT, the machine that unpacks and repacks nucleosomes so RNA polymerase can read through them. Its interest for this project is one specific contact: SSRP1 grips LEDGF/p75, a protein that is itself a documented partner of the leukaemia fusion, which makes SSRP1 exactly two handshakes from the fusion complex even though no paper has ever put them in the same experiment.

The same thing, for a scientist

SSRP1 (Q08945, 709 aa) is the small subunit of the FACT heterodimer [E1]. Its sole annotated domain with residue coordinates is an **HMG box at 547-615** [E5]. Its direct KMT2A literature is a single record and none of the 9 records returned by `"FACT complex" AND (leukemia OR leukaemia)` concerns KMT2A on title-check [E6]. What makes SSRP1 worth a separate entry rather than a line inside the complex node is a specific, mapped, domain-level contact: **LEDGF/p75 complexes with FACT, and the interaction is with SSRP1 specifically, in an hSpt16-independent manner, requiring the PWWP domain of LEDGF and the HMG domain of SSRP1** [E2]. LEDGF/p75 is PSIP1, which this graph already carries as a KMT2A-fusion interactor: its integrase-binding domain is the module that tethers the menin-MLL complex to chromatin. The FACT-side contact is made by a different LEDGF module (PWWP, the H3K36me-reading domain) than the one that binds menin-MLL (IBD), so the two contacts are **not mutually exclusive by construction** and a ternary arrangement is geometrically permitted. **No experiment has tested it.** Dependency is a dead end in the usual way and slightly more extreme than its partner subunit: pan-essentiality -1.2714 across 1,208 lines with a KMT2A-rearranged difference of **+0.2651, rank 17,672 of 17,728 (99.68th percentile)** - the 57th most anti-selective gene in the genome by this statistic [E3] - and the lineage split shows no requirement in either half (myeloid raw -1.233, z +0.54; B-lineage raw -1.118, z +0.90) [E4].

What it looks like
FACT complex subunit SSRP1709 aa
HMG boxHMG box · residues 547–6150200400600
loading AF-Q08945-F1…
AlphaFold AF-Q08945-F1 ↗
colourshape

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

very high (90+)confident (70–90)low (50–70)very low (<50)
I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper
The LEDGF/p75 bridge, stated precisely

Lopez and colleagues used protein-protein interaction assays to show that LEDGF/p75 complexes with FACT, and then took the interaction apart: the contact is with SSRP1, not with hSpt16, it is hSpt16-independent, and it requires the PWWP domain of LEDGF proteins and the HMG domain of SSRP1 2. They also showed a functional consequence unrelated to this project - partial SSRP1 knockdown reduces HIV-1 infection in human CD4+ T cells in an LEDGF-independent manner, affecting LTR-driven expression rather than integration.

For this graph the relevant chain is:

  • gene-psip1 (LEDGF/p75) is a documented component of the menin-MLL chromatin-tethering module.
  • LEDGF's IBD is the module that engages menin-MLL.
  • LEDGF's PWWP is the module that engages SSRP1 2 - and PWWP is also the H3K36me2/3-reading module that gives LEDGF its gene-body preference.

So the two contacts use different domains of LEDGF, which is the reason this is worth writing down: a competition model is not forced by the architecture. The honest statement is that SSRP1 is two mapped contacts from KMT2A::AFF1, that nobody has looked, and that the graph should record the geometry rather than the conclusion.

Counter-consideration that has to be stated in the same breath. The system is HIV-1 biology in CD4+ T cells and 293T-type interaction assays. LEDGF/p75 is a famously promiscuous chromatin adaptor with a long list of PWWP- and IBD-mediated partners, and the fact that an interactome switches with phosphorylation state is documented for exactly this protein (PMID 29997176, one of the five hits in this beat's MED1 query). A mapped domain-level contact is much better evidence than a mass-spec co-occurrence, but it is not evidence that the contact happens in a leukaemia blast.

Cell type, cell state, subcellular and subnuclear localisation, epigenetics

Subcellular: nucleus and chromosome. Subnuclear: nucleolus, which UniProt annotates for SSRP1 and does not annotate for its obligate partner SPT16 5. That asymmetry within an obligate heterodimer is worth flagging: either SSRP1 has a FACT-independent nucleolar pool, or the annotations are simply of different vintage. This beat did not resolve it, and it is recorded as unresolved rather than smoothed over.

Cell type: in this disease, only MLL-r leukaemia cell lines and MLL-r ALL patient-derived xenograft cells, and only as the target of CBL0137 7. In the LEDGF work, human CD4+ T cells 2. Cell state: not established in any KMT2A context.

Epigenetic readout: none written directly - FACT is a chaperone, not a writer. The nearest epigenetic quantity is nucleosome occupancy / H2A-H2B exchange 1. Indirectly, the SSRP1 partner LEDGF reads H3K36me2/3 through the same PWWP domain that binds SSRP1 2, so if the ternary arrangement exists it would couple FACT to an H3K36-marked compartment - which is a prediction, not a finding.

Why the dependency number here is worth almost nothing

SSRP1 ranks 17,672 of 17,728 on the KMT2A-selectivity statistic 3. Read naively that says KMT2A-rearranged lines are protected by needing SSRP1 less. That reading is wrong for the same reason the corresponding FACT number is wrong: at a pan-essentiality of -1.27 every arm is dying, and a difference between two lethal effects is dominated by where each arm happens to sit against the Chronos floor.

The lineage split makes the point sharper rather than resolving it. Both halves score positive (myeloid z +0.54 on raw -1.233; B-lineage z +0.90 on raw -1.118) against an other-leukaemia mean of -1.404 4. A gene whose two index halves agree is normally the good case - here they agree on 'no selective requirement', and the raw effects show why the z looks dramatic: the arms differ by about 0.2 Chronos units in a regime where everything is below -1.1.

This is the rule-9 case in miniature. The z is scaled by the comparator SD (0.317), which is small because every leukaemia line needs FACT, so a modest raw difference becomes a large z. Quoting +0.90 without the raw -1.118 beside it would misdescribe the biology.

Correction: there IS a KMT2A paper about FACT, on the index fusion, and we cite it elsewhere

6 records that "FACT complex" AND (leukemia OR leukaemia) returns 9 records "of which title-checking finds none about KMT2A". The queries are accurate and the conclusion is wrong.

PMID 37626123 - MLL-AF4 cooperates with PAF1 and FACT to drive high-density enhancer interactions in leukemia (Crump ... Milne, Nat Commun 2023). MLL-AF4 is the index fusion. MLL-AF4 target enhancers are enriched for PAF1C and FACT, and loss of these factors abolishes enhancer-promoter contact. This graph cites that paper in four other entries.

It is invisible to this entry's queries because it says FACT, not FACT complex, and never names a subunit. Verified 2026-09-20: (SSRP1 OR SUPT16H) AND (MLL OR KMT2A) returns 1 record and it is not this paper; "MLL-AF4" AND FACT returns 5 and it is first.

What changes and what does not. The claim that SSRP1 has no KMT2A literature is withdrawn. The dependency picture is untouched - SSRP1 remains less required in KMT2A-rearranged lines (difference +0.265, permutation p 0.0104 at pan-essentiality -1.271), and that is not in conflict: a factor can be required for an enhancer-promoter architecture and required no more than everywhere else for survival, and a knockout screen measures survival. A structural requirement with no differential survival cost is exactly the shape of a target a viability screen cannot find.

Facts and figures
uniprotQ08945
gene symbolSSRP1
protein nameFACT complex subunit SSRP1
length AA709
alphafold IDAF-Q08945-F1
accession resolutionUniProt REST query gene_exact:SSRP1 AND organism_id:9606 AND reviewed:true, run 2026-09-19; one reviewed hit; primary gene name read back as SSRP1 and length 709 confirmed from the same fetched record.
complexFACT (FAcilitates Chromatin Transcription): SPT16 + SSRP1 heterodimer
domains noteUniProt annotates exactly one feature with coordinates, the HMG box 547-615, which is also the module mapped as binding the LEDGF/p75 PWWP domain. The pleckstrin-homology-like and middle domains described structurally have no UniProt coordinates and are omitted rather than estimated.
functionsmall subunit of the FACT histone chaperone; HMG-box DNA/nucleosome binding
classificationNOT a documented interactor of KMT2A::AFF1; two mapped contacts away via LEDGF/p75 (PSIP1)
interaction evidence classbiochemical purification (FACT); protein-protein interaction assays with domain mapping (LEDGF PWWP to SSRP1 HMG)
shown for fusionnone
cell type evidenceMLL-r leukaemia lines and MLL-r ALL PDX cells (drug only); human CD4+ T cells (LEDGF work)
cell state evidencenot established
epigenetic readoutnone written; nucleosome occupancy and H2A-H2B exchange. Indirect link to H3K36me2/3 via the LEDGF PWWP domain that binds the SSRP1 HMG box.
enumeration record counts
(MLL OR KMT2A) AND (SSRP1…1.000(MLL OR KMT2A) AND (SSRP1 OR T160 OR FACT80): 1(MLL OR KMT2A) AND SSRP11.000(MLL OR KMT2A) AND SSRP1: 1"FACT complex" AND (leuke…9.000"FACT complex" AND (leukemia OR leukaemia): 9
PDB IDS
[
  "4IFS",
  "5UMR",
  "5UMS",
  "5VWE",
  "6L1E",
  "6L1R",
  "6L34",
  "6UPK",
  "6UPL",
  "9EH2",
  "9S3G"
]
aliases retired or alternate
[
  "FACT80",
  "T160"
]
complex partners
[
  "gene-supt16h",
  "protein-fact-complex"
]
domains
[
  {
    "name": "HMG box",
    "start": 547,
    "end": 615
  }
]
mapped interfaces
[
  {
    "partner": "gene-psip1",
    "ssrp1_module": "HMG box (aa 547-615)",
    "partner_module": "PWWP domain of LEDGF/p75",
    "spt16_dependent": false,
    "system": "HIV-1 / CD4+ T cells",
    "evidence": "E2"
  }
]
localization
{
  "subcellular": "nucleus; chromosome (UniProt)",
  "subnuclear": "nucleolus (UniProt) - annotated for SSRP1 but not for its obligate partner SPT16; unresolved asymmetry. No subnuclear data in any KMT2A system."
}
essentiality
{
  "depmap_scan": "analyses/depmap-kmt2a-selective-scan-full",
  "pan_essentiality_all_lines": -1.2714,
  "n_all_lines": 1208,
  "mean_kmt2a_rearranged": -1.1468,
  "mean_other_leukemia": -1.412,
  "kmt2a_selectivity_difference": 0.2651,
  "rank": 17672,
  "n_scanned": 17728,
  "percentile": 99.684,
  "lineage_split": {
    "analysis": "analyses/depmap-interactome5-lineage-split",
    "raw_mean_kmt2a_myeloid": -1.233,
    "z_kmt2a_myeloid": 0.54,
    "raw_mean_kmt2a_b_lineage": -1.118,
    "z_kmt2a_b_lineage": 0.9,
    "raw_mean_other_leukaemia": -1.404,
    "sd_other_leukaemia": 0.317,
    "raw_effect_OCI_AML3": -1.566,
    "z_OCI_AML3": -0.51,
    "n_kmt2a_myeloid_lines": 6,
    "n_kmt2a_b_lineage_lines": 4,
    "n_other_leukaemia_lines": 45,
    "verdict": "no requirement in either lineage half; both halves score positive"
  },
  "interpretation": "pan-essential and the most anti-selective gene in this beat's panel; the sign is a floor artefact, not protection"
}
enumeration record counts — the numbers behind the chart
[
  {
    "query": "(MLL OR KMT2A) AND (SSRP1 OR T160 OR FACT80)",
    "records": 1
  },
  {
    "query": "(MLL OR KMT2A) AND SSRP1",
    "records": 1
  },
  {
    "query": "\"FACT complex\" AND (leukemia OR leukaemia)",
    "records": 9,
    "records_after_title_check": 0
  },
  {
    "date": "2026-09-19"
  }
]
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -1.1468,
  "chronos_mean_other_leukaemia": -1.412,
  "difference": 0.2651,
  "pan_essentiality_all_lines": -1.2714,
  "selectivity_rank": 17672,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
How this could bite us back
  • SSRP1 is pan-essential at -1.27 across 1,208 lines with no KMT2A window at all. Targeting it is targeting transcription in every dividing cell.
  • The LEDGF bridge, if it turned out to be real, would argue for the opposite of an intervention: LEDGF/p75 is already the tethering module the menin-MLL axis depends on, and adding a second essential chaperone to that module widens rather than narrows the toxicity surface.
  • An SSRP1-directed agent would be indistinguishable in effect from a FACT-directed agent, because the subunits are obligate partners. There is no subunit-selective handle here.
  • CBL0137 data must not be read as validating an SSRP1 loss-of-function strategy: the compound traps FACT on chromatin and the genetic data says removing FACT gives no KMT2A-selective benefit.
Still open
  1. Does the LEDGF PWWP-SSRP1 HMG contact occur in a KMT2A-rearranged blast, and can LEDGF bridge FACT to the menin-MLL module simultaneously? The domains used are different (PWWP for SSRP1, IBD for menin-MLL), so the ternary complex is geometrically allowed and has simply never been tested.
  2. Why is SSRP1 annotated nucleolar when its obligate partner SPT16 is not? Either there is a FACT-independent SSRP1 pool or the annotations differ in vintage; a simple fractionation would say which.
  3. Does SSRP1 depletion change H3K36me2/3 distribution, given that its mapped partner domain on LEDGF is the H3K36me reader? Nothing read here tests it.
  4. Is the SSRP1 anti-dependency reproducible in an expression-normalised or line-normalised analysis, or does it disappear when floor compression is handled properly? The repository already holds scripts that would answer this and this beat did not run them.
Every source behind this entry (8)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    FACT comprises a human homologue of S. cerevisiae Spt16/Cdc68 together with the high-mobility-group-1-like protein structure-specific recognition protein 1 (SSRP1); FACT specifically interacts with nucleosomes and H2A/H2B dimers and its activity is abrogated by crosslinking nucleosomal histones.

    The chromatin-specific transcription elongation factor FACT comprises human SPT16 and SSRP1 proteins · Orphanides G, Wu WH, Lane WS, Hampsey M, Reinberg D · Nature 400(6741):284-8 1999
    PMID 10421373DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, all identifiers copied from that fetched record. Purely in vitro reconstitution; no cellular, haematopoietic or KMT2A context. This citation establishes SSRP1's membership of FACT and nothing about leukaemia.

    Evidence that argues against this (1)
    • SSRP1 is annotated to the nucleolus while SPT16 is not, which is difficult to reconcile with a strictly obligate heterodimer and leaves open a FACT-independent SSRP1 pool that this 1999 biochemistry would not have detected. UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    LEDGF/p75 complexes with the FACT heterodimer; the interaction is with SSRP1 in an hSpt16-independent manner and requires the PWWP domain of LEDGF proteins and the HMG domain of SSRP1. Partial shRNA knockdown of SSRP1 reduces HIV-1 infection but not Murine Leukemia Virus infection in human CD4+ T cells, affecting LTR-driven gene expression rather than viral DNA integration.

    The Structure-Specific Recognition Protein 1 Associates with Lens Epithelium-Derived Growth Factor Proteins and Modulates HIV-1 Replication · Lopez AP, Kugelman JR, Garcia-Rivera J, Urias E, Salinas SA, Fernandez-Zapico ME, Llano M · Journal of Molecular Biology 428(14):2814-31 2016
    PMID 27216501DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, and every identifier copied from that fetched record. **The domain mapping (PWWP to HMG) is stated in the abstract; the assay format, the direction of the pulldown and whether the contact is direct were not verified from full text.** The system is HIV-1 biology in CD4+ T cells; there is no leukaemia, no KMT2A and no fusion protein anywhere in this work. LEDGF/p75 is a promiscuous chromatin adaptor, so a mapped contact is not by itself evidence of a physiologically dominant complex.

    Evidence that argues against this (1)
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    In the project's genome-wide KMT2A-rearranged selectivity scan of 17,728 genes, SSRP1 has a pan-essentiality of -1.2714 across 1,208 lines and a KMT2A-rearranged selectivity difference of +0.2651 (KMT2A-rearranged mean -1.1468 against other-leukaemia mean -1.412), ranking 17,672 of 17,728 (99.684th percentile). Its partner SUPT16H is +0.1165 at rank 17,162.

    Genome-wide KMT2A-rearranged selectivity scan, 17,728 genes (DepMap Chronos via Open Targets) · Broad Institute DepMap / Open Targets Platform; analyses/depmap-kmt2a-selective-scan-full 2026
    What this evidence cannot show

    Rows read out of analyses/depmap-kmt2a-selective-scan-full/results.json on 2026-09-19 in a scripted pass; no number typed from recall. Eleven KMT2A-rearranged against 43 other leukaemia lines, AML-heavy, none infant-derived. **Severe floor compression**: at a pan-essentiality of -1.27 the difference statistic is a difference between two lethal effects and its sign is not reliable evidence of protection.

    Evidence that argues against this (1)
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    On the lineage-split dependency run for this beat, SSRP1 shows no KMT2A-rearranged requirement in either lineage half: KMT2A-rearranged myeloid raw mean gene effect -1.233 (z +0.54 in other-leukaemia SDs), B-lineage raw mean -1.118 (z +0.90), against an other-leukaemia mean of -1.404 with SD 0.317 and a non-leukaemia mean of -1.267. OCI-AML3 (NPM1c) sits at raw -1.566 (z -0.51) and U-937 at z -0.11.

    Interactome round 5 panel, lineage-split DepMap dependency (FACT, NSL, MSL, COMPASS remainder, Mediator) · Broad Institute DepMap / Open Targets Platform; analyses/depmap-interactome5-lineage-split 2026
    What this evidence cannot show

    Rows read out of analyses/depmap-interactome5-lineage-split/results.json on 2026-09-19. Six myeloid and four B-lineage index lines, so one line moves a mean; the split is by lineage LABEL, not expression signature. Raw effects are quoted beside each z because the comparator SD is small (0.317) precisely because every leukaemia line requires FACT, which inflates z relative to the raw difference.

    Evidence that argues against this (1)
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProtKB Q08945 (SSRP1_HUMAN, FACT complex subunit SSRP1) is 709 aa with the synonym FACT80, is annotated to nucleus, nucleolus and chromosome, has 11 PDB cross-references, and carries one annotated feature with residue coordinates: an HMG box DNA-binding region at residues 547-615.

    UniProtKB Q08945 (SSRP1_HUMAN), FACT complex subunit SSRP1 · UniProt Knowledgebase, reviewed (Swiss-Prot) 2026
    What this evidence cannot show

    Record fetched 2026-09-19 via UniProt REST with gene_exact:SSRP1 AND organism_id:9606 AND reviewed:true; **primary gene name read back as SSRP1 and length 709 confirmed from the same fetched record**, per the standing accession rule. One reviewed hit, so no synonym collision to disambiguate - which is worth stating because SSRP1 was flagged in this beat's brief as short and collision-prone. The pleckstrin-homology-like and middle domains described in the structural literature have no UniProt coordinates and are deliberately omitted rather than estimated.

    Evidence that argues against this (1)
    • The alias T160 is a phosphosite label used widely in the kinase literature and the alias FACT80 is rare, so an alias-broadened PubMed sweep on this gene inflates rather than improves recall. This is the precision half of the standing alias rule and it applied here. NCBI PubMed, queried 2026-09-19 2026
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    As of 2026-09-19, PubMed returns 1 record for '(MLL OR KMT2A) AND (SSRP1 OR T160 OR FACT80)' and 1 for '(MLL OR KMT2A) AND SSRP1'. The broader '"FACT complex" AND (leukemia OR leukaemia)' returns 9 records, of which title-checking finds none about KMT2A - they concern avian leukosis virus integration, curaxins in JAK2 V617F and BCR::ABL1 disease, an ENL super-enhancer study, Daxx/SWI-SNF heterochromatin, a 1988 Philadelphia-translocation cytogenetics paper, the LEDGF/SSRP1 HIV study, and two metal-complex cytotoxicity papers.

    What this evidence cannot show

    Counts from the esearch `count` field, queried 2026-09-19, and **every record in the 9-record query was title-checked individually** rather than counted. Symbol set searched: SSRP1, T160, FACT80, plus the phrase 'FACT complex'. Not searched: 'structure-specific recognition protein' as a phrase, or the retired-style name 'recombination signal sequence recognition protein'. A counted absence is a floor on recall.

    Evidence that argues against this (1)
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The curaxin CBL0137 acts by trapping FACT into chromatin, and is active against 12 MLL-rearranged leukaemia cell lines, 3 MLL-r ALL patient-derived xenograft cell populations, a subcutaneous MV4;11 MLL-r AML model and 5 MLL-r ALL patient-derived xenograft models, through p53 pathway activation and an interferon response.

    Potent antileukemic activity of curaxin CBL0137 against MLL-rearranged leukemia · Somers K, Kosciolek A, Bongers A, El-Ayoubi A, Karsa M, Mayoh C, Wadham C, Middlemiss S, Neznanov N, Kees UR, Lock RB, Gudkov A, Sutton R, Gurova K, Haber M, Norris MD, Henderson MJ · International Journal of Cancer 146(7):1902-1916 2020
    PMID 31325323DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read, identifiers copied from the fetched record. The compound targets the FACT heterodimer and this entry does not claim it is SSRP1-selective. The mechanism is p53- and interferon-dependent rather than KMT2A-directed, and there is no non-MLL-r leukaemia comparator arm in the abstract.

    Evidence that argues against this (1)
  8. E-fact-enhancerestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.journal-article

    MLL-AF4 binding promotes broad, high-density regions of enhancer-promoter interactions at a subset of key targets; those enhancers are enriched for transcription elongation factors including PAF1C and FACT; and loss of these factors abolishes enhancer-promoter contact.

    PMID 37626123DOI
    What this evidence cannot show

    Record fetched 2026-09-20, title/journal/year/DOI/first and last author read back; claim from the abstract, **full text not read here**. Established as what the paper reports, in MLL-AF4 leukaemia models by Micro-Capture-C. **Enrichment and loss of contact on depletion is not a survival requirement** and is not evidence for a FACT-directed therapy on its own.

    Evidence that argues against this (1)
    • This entry declared the KMT2A-FACT literature empty while the graph cited this paper in four other entries. The defect was not the query alone - nothing here checks an absence claim against the graph's own bibliography, which is the cheaper fix and is now an open question at `ana-a-subunit-keyed-absence-is-structurally-wrong`. Oncofusion knowledge graph (internal analysis) 2026
Connected to