Oncofusion

start / graph / gene-fanca

Proteinmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q-fa-replication

FANCA — FA core complex; rank 20 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all

also called FANCA · FAA · FACA · FANCH · O15360

FANCA is part of a cellular repair crew that fixes DNA when the copying machinery gets stuck. A screen run by this project found that leukaemia cells carrying the KMT2A fusion need it more than other leukaemia cells do — and nobody has ever published a paper on this gene in KMT2A leukaemia, so the finding is genuinely new and genuinely unconfirmed.

The same thing, for a scientist

FANCA (UniProt O15360, 1455 aa, primary gene name asserted against the accession) sits in the FA core complex — the largest subunit, in the FANCA-FANCG-FAAP20 arm. It entered this graph through a DepMap sweep in which the 26-gene FA/BRCA annotation beats a permutation null at p = 5e-05, FANCA ranking 20 of 8,269 by raw effect (d = -0.296) [E2]. Re-testing here: the depletion is present in BOTH lineage halves (-0.2995 myeloid, -0.4565 lymphoid) and stronger in the lymphoid one; it survives a line-label permutation that preserves subunit correlations (p = 0.0002); and FANCA is not pan-essential (mean Chronos -0.2588 across all screened lines) [E3][E4]. Three things cut against it. Non-KMT2A B-ALL lines show about half the same effect, so a large part of this is B-lymphoid lineage rather than KMT2A rearrangement [E3]. The obligate FA core subunits do NOT move together, which is hard to reconcile with a canonical pathway dependency. And Fanconi anaemia patients, who have no FA pathway at all, nonetheless develop KMT2A-rearranged leukaemia [E7].

Go deeper
What the protein does, and which part of the pathway it sits in

Module: FA core complex — the largest subunit, in the FANCA-FANCG-FAAP20 arm.

UniProt describes FANCA as a DNA repair protein that may act in postreplication repair or a cell-cycle checkpoint, implicated in interstrand cross-link repair, and places it in the multisubunit FA core complex 1.

Why the module matters for reading the DepMap result: the FA core complex is an obligate multi-subunit ubiquitin ligase whose only well-established job is to monoubiquitinate the FANCD2-FANCI (ID2) heterodimer. If the dependency seen here were the canonical pathway working as a unit, every obligate core subunit and the ID2 substrate should move together. They do not — see mech-fa-core-complex-heterogeneity-kmt2a. That is the single most important caveat on this entry and it is not a small one.

How this gene entered the graph, and the numbers

FANCA was not found in the literature. It was found by a scan, and then by a pathway permutation over a gene set fixed before any result was read 2.

statisticvalue
rank by raw effect, 8,269-gene sweep20
rank by scaled z9
d (Chronos difference), whole index arm-0.296
d, myeloid half-0.2995
d, lymphoid half-0.4565
z, myeloid half-1.342
z, lymphoid half-2.045
pan-cancer Chronos, all screened lines-0.2588
fraction of all lines below -0.514.2%
line-label permutation p0.0002

Two of those rows carry most of the weight. The lineage split 3 is the one the project asked for, and the answer is not the one the myeloid-artefact worry predicted: this gene is depleted MORE in the four B-lineage lines than in the seven myeloid ones. The line-label permutation 3 is the answer to the complex-correlation objection: redrawing which lines are the index arm, with the gene set held fixed and the 7-myeloid/4-lymphoid composition reproduced, preserves every correlation between subunits, and the observed depletion still sits at p = 0.0002.

And FANCA is not pan-essential 4, which is the filter that removed most of the earlier scan's contaminants.

The published literature, and a counted absence

On FANCA in KMT2A-rearranged leukaemia there is nothing. 5 The exact query and the title-checked count are in the evidence item; aliases searched were FANCA, FAA, FACA, FANCH.

On FANCA in leukaemia at all, 66 records. 6

And the counter-direction. Human FA patients develop KMT2A-rearranged leukaemia 7. A cell that has lost the FA pathway entirely can still acquire a KMT2A fusion and expand into a clinical leukaemia, which rules out the strongest reading of the DepMap result — that KMT2A-rearranged cells REQUIRE this pathway — and leaves only the quantitative one.

Regulated nuclear import, and a genuine subnuclear absence (Q87, 2026-09-21)

Subcellular. Major form nuclear, minor form cytoplasmic (UniProt O15360 - both terms UNCODED). The split is mechanistic: FANCA has an N-terminal NLS, FANCG's N-terminal two-thirds binds it, and complex formation is coupled to nuclear translocation and to function (PMID 10961856). HPA v25.1 IF: nucleoplasm, Supported.

Subnuclear: BOUNDED ABSENCE (measured-not-reported). No named compartment - not nucleolus, PML body, speckle, lamina or heterochromatin. What exists instead is damage-induced chromatin recruitment, usually read out through FANCD2 foci rather than FANCA's own.

Cell type. Haematopoietic stem and progenitor cells, by human genetics - Fanconi anaemia is a marrow-failure syndrome - not by any imaging experiment in a blood cell.

Cell state. Replication stress and interstrand-crosslink repair; this graph already holds cell-state-replication-stress-icl-repair.

Shown for fusion. No, in the strong sense, and consistent with this entry's existing finding that FANCA ranks 20th of 8,269 for KMT2A-selective dependency with no KMT2A literature at all.

Facts and figures
uniprotO15360
length AA1455
uniprot resolutiongene_exact:FANCA AND organism_id:9606 AND reviewed:true; primary gene name read back as FANCA; length 1455 aa checked against the entry
fa moduleFA core complex — the largest subunit, in the FANCA-FANCG-FAAP20 arm
subnuclear localizationNO NAMED COMPARTMENT - bounded absence, class measured-not-reported. FANCA has a regulated nuclear/cytoplasmic split (major nuclear, minor cytoplasmic; UniProt O15360, both terms uncoded) driven by an N-terminal NLS to which FANCG binds (PMID 10961856), but no paper assigns it to a nucleolus, PML body, speckle, lamina or heterochromatin compartment. The available sub-nuclear information is damage-induced recruitment to repair foci, and even that is usually reported through FANCD2 foci rather than FANCA's own.
cell stateREPLICATION STRESS / INTERSTRAND-CROSSLINK REPAIR. FANCA's activity is conditional on damage: the FA core complex acts when replication forks meet interstrand crosslinks, and this graph already carries a cell-state-replication-stress-icl-repair node for exactly this state. The compartment consequence of that state - where FANCA goes when a fork stalls - is the part nobody has imaged directly.
shown for fusionNO - nobody has looked in any fusion, in either sense. Measured 2026-09-21: FANCA[tiab] AND the eight-term fusion clause AND the ten-term localisation clause returns 0 records. This entry's own title already records that FANCA ranks 20th of 8,269 for KMT2A-selective dependency with NO KMT2A literature at all, so the localisation absence sits inside a wider one and is not a separate surprise.
essentiality
{
  "depmap_common_essential": false,
  "chronos_mean": -0.2588,
  "frac_lines_below_minus_0_5": 0.142
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.6377,
  "chronos_mean_other_leukaemia": -0.2786,
  "difference": -0.3591,
  "pan_essentiality_all_lines": -0.2588,
  "selectivity_rank": 56,
  "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."
}
literature
{
  "kmt2a_records_raw": 5,
  "kmt2a_records_title_checked": 1,
  "leukaemia_records": 66,
  "aliases_searched": [
    "FANCA",
    "FAA",
    "FACA",
    "FANCH"
  ],
  "kmt2a_query": "(KMT2A[tiab] OR MLL[tiab] OR \"mixed-lineage leukemia\"[tiab]) AND (FANCF[tiab] OR FANCA[tiab] OR FAA[tiab] OR FACA[tiab] OR FANCH[tiab] OR FANCG[tiab] OR XRCC9[tiab])",
  "leukaemia_query": "(FANCA[tiab] OR FACA[tiab] OR FANCH[tiab]) AND (leukemia[tiab] OR leukaemia[tiab])"
}
localization
{
  "subcellular": "NUCLEUS as the MAJOR form, CYTOPLASM as the MINOR form, and the split is regulated rather than incidental. UniProt O15360 lists Nucleus and Cytoplasm with the note 'The major form is nuclear. The minor form is cytoplasmic'; both terms are UNCODED - no evidence code, no citation - so the database line is not itself a measurement (rule 91). The mechanism behind the split is published: FANCA carries an N-terminal NUCLEAR LOCALIZATION SIGNAL, the amino-terminal two-thirds of FANCG binds directly to that NLS, and complex formation is coupled to nuclear translocation and to functional activity (PMID 10961856). HPA v25.1 IF, reliability Supported, gives a single main location, Nucleoplasm - the major form.",
  "subnuclear": "BOUNDED ABSENCE at the compartment level; a CHROMATIN-RECRUITMENT answer exists instead. No paper places FANCA in a named subnuclear compartment - not the nucleolus, not PML bodies, not speckles, not the lamina. What the field measures instead is damage-induced recruitment: FANCA[tiab] AND (foci[tiab] OR chromatin[tiab]) returns 78 records (2026-09-21, NOT title-checked, quoted as a size only), and the standard readout in that literature is FANCD2 foci rather than FANCA foci - FANCA is assayed as the core-complex subunit whose loss abolishes FANCD2 monoubiquitination. So the honest subnuclear statement is 'damage-induced chromatin/repair foci, measured through a downstream reporter', and the compartment-level answer is genuinely unmeasured.",
  "absence_class_for_the_subnuclear_answer": "measured-not-reported",
  "how_shown": "two-hybrid mapping of the FANCG-FANCA NLS interaction with nuclear-translocation and complementation assays in FA-A and FA-G cells (PMID 10961856); HPA antibody immunofluorescence; UniProt's two location terms are uncoded",
  "node": "localization-chromatin-bound-versus-soluble-fraction-kmt2a"
}
cell type
{
  "haematopoietic_answer": "HAEMATOPOIETIC STEM AND PROGENITOR CELLS, by disease rather than by imaging. Biallelic FANCA loss causes Fanconi anaemia, whose defining phenotype is progressive bone-marrow failure with a strong predisposition to MDS and AML - FANCA[tiab] AND (\"hematopoietic stem\"[tiab] OR \"haematopoietic stem\"[tiab] OR \"bone marrow failure\"[tiab]) returns 229 records (2026-09-21, quoted as a literature size and not title-checked). That is a cell-type answer derived from human genetics, and it is the strongest one available; no FANCA localisation experiment has been done in a blood cell.",
  "expression_atlas": "HPA v25.1: RNA tissue specificity 'Tissue enhanced'; IF reliability Supported (nucleoplasm)."
}
How this could bite us back
  • The FA pathway is a tumour suppressor pathway. Inhibiting it in a leukaemia patient would be expected to raise, not lower, the mutation rate in surviving cells and in normal haematopoiesis — the exact phenotype Fanconi anaemia patients have.
  • Fanconi anaemia is a bone-marrow-failure syndrome. A systemic inhibitor of this pathway carries a named, well-characterised on-target toxicity in the tissue the disease lives in.
  • Children cured of leukaemia live for decades. A drug that disables interstrand cross-link repair sits badly beside the secondary-malignancy risk that already follows topoisomerase-II-directed therapy for KMT2A-rearranged disease.
  • FANCA has no reported small-molecule ligand, and only UBE2T among the eight carries a catalytic site of the kind ligand discovery usually needs.
  • A quarter-Chronos difference is a narrow therapeutic window even if it is real.
Still open
  1. Is FANCA depleted in KMT2A-rearranged cells because of replication stress at transcribed loci, because of endogenous aldehyde damage as the ALDH2 work in AML argues, or for a third reason?
  2. Does FANCA knockout cause measurable fork slowing or R-loop accumulation in a KMT2A-rearranged line, or only loss of viability?
  3. Is the B-lymphoid component of this signal a property of B-ALL biology, or of the nine non-KMT2A B-ALL lines that happen to be in DepMap?
  4. Do primary KMT2A-rearranged blasts or patient-derived xenografts show the same FANCA requirement, or is it a property of decades-passaged lines?
Every source behind this entry (9)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    FANCA is UniProt O15360, 1455 aa, and the accession's PRIMARY gene name reads back as FANCA. UniProt describes FANCA as a DNA repair protein that may act in postreplication repair or a cell-cycle checkpoint, implicated in interstrand cross-link repair, and places it in the multisubunit FA core complex.

    UniProtKB entry O15360 · UniProtKB/Swiss-Prot 2026
    What this evidence cannot show

    A database record, not an experiment. The accession was resolved by a gene_exact query and then read back: gene_exact matches SYNONYMS, so the primary gene name and the sequence length were both checked against the entry rather than assumed.

    Evidence that argues against this (1)
    • UniProt function text for the FA core subunits is hedged ('may operate in', 'may be implicated in') and largely predates the structural work on the complex; it should not be read as a precise mechanistic claim. UniProtKB/Swiss-Prot 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    In the project's own genome-scale sweep, FANCA ranks 20 of 8,269 by raw effect (d = -0.296) and 9 by the scaled z (-1.329), and the 26-gene FA/BRCA set as a whole beats its permutation null at p = 5e-05 on both statistics.

    Effect-size versus z re-ranking of the genome-scale lineage-consistency sweep, with a permutation test of the Fanconi/FA-BRCA pathway · Oncofusion knowledge graph, analyses/effect-size-vs-z-reranking; DepMap Chronos via Open Targets 2026
    What this evidence cannot show

    DepMap CRISPR gene effect in 11 KMT2A-rearranged leukaemia lines against 43 other leukaemia lines. Eleven immortalised lines, none infant-derived, lineage assigned from DepMap free-text disease strings. A rank is not an effect: d here is a Chronos difference of roughly a quarter to a half, which is real but far from the -0.86 of MEN1.

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

    Splitting the index arm, FANCA is depleted in BOTH halves (d = -0.2995 myeloid, -0.4565 lymphoid; z = -1.342 and -2.045), and the LYMPHOID half is the stronger one. A line-label permutation that holds the gene set fixed and redraws the index arm as 7 myeloid + 4 lymphoid leukaemia lines gives p = 0.0002 for FANCA, and p = 0.0002 for the eight-gene headline set.

    What this evidence cannot show

    The lymphoid half is four cell lines. The line-label permutation is the right null for correlated subunits because it preserves every gene-gene correlation, but it cannot tell a KMT2A effect from anything else those eleven particular lines share.

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

    FANCA is not pan-essential: mean Chronos across every screened cell line is -0.2588, and only 14.2% of lines fall below -0.5. For contrast ACTB is -0.815 with 72.1%.

    What this evidence cannot show

    Computed from the Open Targets slice of DepMap, which does not state its release version. 'Not pan-essential' is a weaker statement than 'selective': a gene can be non-essential everywhere and still not be selectively required in KMT2A-rearranged cells.

    Evidence that argues against this (1)
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    There is no published literature on FANCA in KMT2A-rearranged leukaemia. A PubMed title/abstract search including the retired and alternative symbols FANCA, FAA, FACA, FANCH against KMT2A, MLL and 'mixed-lineage leukemia' returns 5 records, of which 1 is substantive after title-checking every hit. Four of the five are incidental — two cancer mutation-frequency lists (anaplastic thyroid carcinoma, post-transplant plasmablastic lymphoma), one hepatic neuroendocrine exome study, and one 2001 Fanconi mosaicism paper that uses an MLL FISH probe as a cytogenetic reagent. The fifth is substantive.

    What this evidence cannot show

    A counted absence is only as good as its symbol set, and short or word-like aliases inflate counts rather than shrink them, so every hit was title-checked and the checked number is the one reported. Title/abstract search misses work buried in supplementary tables and in screens that name a pathway rather than a gene.

    Evidence that argues against this (1)
    • The pathway is not unstudied even where the gene is: a Hoxa-dependency transcriptome in MLL-rearranged mouse leukaemia returns an FA-pathway-enriched signature, which no gene-symbol query would find. Kettyle LM et al. 2019
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    FANCA in leukaemia more broadly: 66 PubMed records for the symbol and its aliases against leukaemia or leukaemia. The only substantive paper that puts an FA gene and a KMT2A fusion in the same experiment uses Fanca-knockout and MLL-AF9 mouse models in parallel and reports that Trem1 is required for the leukaemogenic activity of pre-leukaemic and leukaemia stem cells in both.

    PMID 35443564DOI
    What this evidence cannot show

    Mouse; the two models are run side by side as separate sources of persistent DNA damage and oncogenic stress.

    Evidence that argues against this (1)
    • The paper does not test whether MLL-AF9 cells need Fanca, which is the question here. Read carefully it is evidence that Fanca-deficiency and MLL-AF9 SHARE a downstream inflammatory programme - which, if anything, frames Fanca loss as a leukaemia-PROMOTING state rather than a dependency. Li X et al. 2022
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Human Fanconi anaemia patients, who are biallelically null for an FA gene, do develop KMT2A-rearranged leukaemia — a documented case of therapy-related AML carrying MLL-CBP t(11;16) arising in a child with Fanconi anaemia after topoisomerase-II-directed chemotherapy for ALL.

    PMID 10679915
    What this evidence cannot show

    A single 2000 case report, and the FA complementation group is not stated in the abstract, so which FA gene was lost is not established from this record. Therapy-related leukaemia in an FA background is also a special case.

    Evidence that argues against this (1)
    • This is the sharpest argument against reading the DepMap signal as an absolute requirement: a KMT2A-rearranged clone arose and expanded in a host with a disabled FA pathway. It does not exclude a quantitative dependency, and the DepMap effect sizes are quantitative, not absolute. Quentin S et al. 2011
  8. E8moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    FANCA's nuclear/cytoplasmic split is regulated by a mapped import mechanism rather than being a passive distribution: FANCA carries an amino-terminal nuclear localization signal, the amino-terminal two-thirds of FANCG (residues 1-428) binds directly to that NLS, and the complex's formation is coupled to nuclear translocation and to functional activity. UniProt O15360 records the resulting distribution as 'The major form is nuclear. The minor form is cytoplasmic'.

    Carboxy terminal region of the Fanconi anemia protein, FANCG/XRCC9, is required for functional activity. · Kuang Y, Garcia-Higuera I, Moran A, Mondoux M, Digweed M, D'Andrea AD · Blood 2000
    PMID 10961856
    What this evidence cannot show

    Blood, 2000, using mutant constructs and two-hybrid analysis rather than endogenous protein, and the functional readout is complementation in FA cell lines. Neither UniProt location term carries an evidence code, so the database line is a curator statement and this paper is the supporting mechanism rather than the annotation's cited source. Nothing haematopoietic in the experiments themselves, despite the disease being a marrow-failure syndrome.

    Evidence that argues against this (1)
    • The regulated-import story argues against the way FANCA is usually treated in dependency work, including in this entry's own DepMap section: a CRISPR knockout removes the protein, but the biology that matters clinically is often a PARTIAL failure of nuclear accumulation caused by a partner defect, since a FANCG mutation that leaves FANCA intact still leaves FANCA in the wrong compartment. That means FANCA's dependency rank and FANCA's pathway state can come apart, and a KMT2A-selectivity difference attributed to FANCA could belong to any member of the core complex whose loss strands it in the cytoplasm. Nothing in the scan can distinguish those. Daddacha W et al. 2017
  9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    No paper assigns FANCA to a named subnuclear compartment, and no FANCA localisation result exists in any KMT2A-fusion context. Measured 2026-09-21: FANCA[tiab] AND (KMT2A[tiab] OR MLL[tiab] OR "MLL-AF4"[tiab] OR "MLL-AF9"[tiab] OR "MLL-ENL"[tiab] OR "MLL-AF10"[tiab] OR "t(4;11)"[tiab] OR "t(9;11)"[tiab]) AND (localization[tiab] OR localisation[tiab] OR subnuclear[tiab] OR nuclear[tiab] OR nucleolar[tiab] OR speckle[tiab] OR speckles[tiab] OR "PML bodies"[tiab] OR chromatin-bound[tiab] OR cytoplasmic[tiab]) returns 0 records. The sub-nuclear information the field does generate is damage-induced recruitment - FANCA[tiab] AND (foci[tiab] OR chromatin[tiab]) returns 78 records - and in that literature the imaged species is usually FANCD2, with FANCA assayed as the core-complex subunit whose loss abolishes FANCD2 monoubiquitination.

    What this evidence cannot show

    Bounded by titles and abstracts (rule 80), and the subnuclear absence is classed measured-not-reported rather than never-measured precisely because a 78-record foci literature exists in which somebody has almost certainly stained FANCA itself. Rule-67 positive control for this query SHAPE: the identical construction, run over all sixteen genes of this tranche on the same day, returned non-zero for seven of them - H2AZ1 1, EED 2, CHAF1B 1, CEBPA 3, TIFAB 1, SAMHD1 1, PROM1 1 - so the instrument can return a one. The querytranslation was read back on every run and confirms the quoted cytogenetic terms expand to "t 4 11"[Title/Abstract] and "t 9 11"[Title/Abstract] and not to a UID filter (rule 95); every term carries its own [tiab] tag rather than sitting inside a parenthesised group (rule 94).

    Evidence that argues against this (1)
    • Calling this an absence risks implying nothing is known about where FANCA acts, and the DNA-damage literature says otherwise in a way that is directly relevant to this graph: SAMHD1, another gene in this same tranche, is recruited to DNA double-strand breaks and complexes with CtIP to promote end resection, and the assay that shows it - recruitment to laser or endonuclease-induced breaks - is a standard one that would work unchanged for FANCA. The reason FANCA has no compartment record is not that the question is hard but that nobody has framed FANCA's address as the result; it is always the pathway's output that gets reported. Daddacha W et al. 2017
Connected to