Oncofusion

start / graph / gene-gata2

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q63-hoxa-cofactors

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

also called NFE1B · GATA-binding protein 2 · Endothelial transcription factor GATA-2 · GATA-2

GATA2 is a master switch for blood stem cells, and leukaemias that keep it switched on are the ones that resist chemotherapy. But in cell-line experiments it is the single least important gene for KMT2A-fusion leukaemia out of nearly eighteen thousand tested, so whatever it is doing, it is not being a partner of the fusion.

The same thing, for a scientist

GATA2 (P23769, 480 aa, 3q21.3) is a heptad member [E1] and a master haematopoietic stem-cell regulator. Its strongest leukaemia result is a **state** result: in a murine **KMT2A-MLL3-driven AML**, leukaemia originating from a stem cell or immature progenitor has higher Gata2 and is more resistant to doxorubicin; Gata2 deletion produces a more robust p53 induction after doxorubicin; GATA2 transcriptionally regulates **RASSF4**, a modulator of the p53 inhibitor MDM2; and doxorubicin plus nutlin-3 is synergistic in Gata2-high murine AML and in patient samples [E2]. In B-ALL, aberrant GATA2 activation is rare and specific: of 629 St Jude Cloud transcriptomes, 13 were GATA2 outliers, **two of which were the only two KMT2A::USP2 cases in the cohort**, with a myeloid-like differential signature and BAG3/EPOR among the targets; a separate patient had *cis* activation via a somatic focal deletion disrupting a TAD boundary near GATA2 [E3]. Against that, GATA2 also behaves as a **haploinsufficient tumour suppressor**: inv(3)/t(3;3) AML repositions a distal GATA2 enhancer to activate EVI1 and simultaneously renders GATA2 functionally haploinsufficient - the same lesion that causes familial AML/MDS and MonoMac/Emberger syndromes [E4]. **And the dependency data are emphatic in the negative direction.** In this project's genome-wide scan GATA2 ranks **17,591 of 17,728** on KMT2A-selectivity with a difference of **+0.1997** - it is among the most anti-selective genes in the entire genome, more required in non-KMT2A leukaemia lines (-0.166) than in KMT2A-rearranged ones (+0.033) - and a live lineage-split re-query gives z **+0.49** myeloid and **+0.75** B-lineage, positive in both halves [E5]. Title-checking the 20 inspected of 21 records returned by `GATA2[tiab] AND (MLL OR KMT2A)[tiab]` found exactly **two on-topic papers - [E2] and [E3] - and eighteen mutational-landscape, CEBPA-subgroup and unrelated cohort papers** [E6].

What it looks like
GATA2480 aa
GATAGATA-type zinc finger 1 (N-finger) · residues 295–319GATAGATA-type zinc finger 2 (C-finger) · residues 349–3730100200300400
loading AF-P23769-F1…
AlphaFold AF-P23769-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 negative result, stated plainly because it is the finding

Of the six genes this beat was sent to write, GATA2 has the weakest case relative to its mention count, and the evidence for that is quantitative rather than impressionistic:

measureGATA2
genome-wide KMT2A-selectivity rank17,591 of 17,728 (99.2nd percentile, anti-selective)
KMT2A-r mean Chronos+0.033 (other leukaemia lines: -0.166)
lineage split z, myeloid / B-lineage+0.49 / +0.75 - positive in BOTH halves
on-topic papers among 20 title-checked of 212: one murine KMT2A-MLL3 chemoresistance model and one two-patient expression correlation
node files in this graph naming GATA211, of which the majority are about inv(3)/GATA2-enhancer hijacking of EVI1

A gene that is more required in the comparator arm than in the index arm, in both lineage halves, with a positive mean effect in the index arm, is not a dependency of this disease by any reading of a CRISPR screen. The honest summary is that GATA2 reached this beat's list because the graph discusses it, and the graph discusses it mostly in the context of a different lesion.

What the negative does NOT bound. A fitness screen cannot see a chemoresistance phenotype, and 2's entire claim is about drug response under doxorubicin - a condition DepMap does not run. It also cannot see a two-patient subtype effect. So GATA2 is not a KMT2A-fusion fitness dependency and may still be a KMT2A-fusion chemoresistance modifier, which is a different and clinically larger question.

Expressed in / required in / maintains a state
  • Expressed in. Restricted in normal tissue to HSCs, haematopoietic progenitors, early erythroid and megakaryocytic cells 3. In leukaemia, substantial intra- and inter-patient variability across AML samples, varying by molecular subtype 2. In B-ALL it is normally OFF, and aberrant activation occurs in about 2% of cases (13 of 629) 3.
  • Required in. Nowhere in KMT2A-rearranged leukaemia that has been demonstrated. Not a CRISPR fitness dependency in any leukaemia arm, and anti-selective in the index arm 5. GATA2 is essential for normal HSC generation and survival 3, which is a requirement in the normal system and a toxicity statement rather than a target statement.
  • Maintains a state. This is where the evidence is. GATA2 maintains a stemness-associated, chemoresistant state: in murine KMT2A-MLL3 AML, stem-cell-origin leukaemia is Gata2-high and doxorubicin-resistant, and the mechanism runs GATA2 -> RASSF4 -> MDM2 -> dampened p53 2. In the two KMT2A::USP2 B-ALL cases, GATA2 activation associates with a myeloid-like transcriptional signature in a B-lineage blast 3 - a state claim, and one that fits this graph's mechanism-kmt2a-mpal-mixed-phenotype and mechanism-lineage-switch-kmt2a.
The two opposite signs, which must both be carried

GATA2 is one of the genes where the leukaemia literature runs in both directions and an entry that reports only one of them misleads.

  • Too much is bad. GATA2-high leukaemia is chemoresistant 2; aberrant GATA2 activation in B-ALL associates with a myeloid-like signature and tumour-associated targets 3.
  • Too little is also bad. GATA2 functional haploinsufficiency is the cause of sporadic and familial AML/MDS and of MonoMac and Emberger syndromes, and inv(3)/t(3;3) AML produces exactly that haploinsufficiency as a side effect of stealing GATA2's enhancer for EVI1 4.

So any intervention lowering GATA2 is moving toward a known leukaemia-predisposition genotype. That is unusual - most oncogenic transcription factors do not have a named germline haploinsufficiency syndrome at the other end - and it belongs in the bite-back rather than in a caveat.

The KMT2A::USP2 result and what it can carry

Two of 629 B-ALL transcriptomes were GATA2 outliers carrying KMT2A::USP2, and they were the only two KMT2A::USP2 cases in the cohort - a 2/2 hit rate, with other KMT2A-rearranged patients showing no GATA2 outlier expression despite high wild-type USP2 in some of them, which the authors read as evidence that the chimeric protein rather than USP2 dosage drives GATA2 3.

n = 2. This graph's existing mech-kmt2a-usp2-gata2-trans-activation already grades it weak and states that it is the entire published transcriptional characterisation of that fusion, with no ChIP-seq, no CUT&RUN, no histone-mark profile and no HOXA/MEIS1 measurement. That grading is correct and this entry does not raise it. The observation is worth carrying because a 2/2 rate in a 629-patient cohort is a striking coincidence, and because it is the only place in the literature where a KMT2A fusion and GATA2 activation are connected causally at all.

Identifier notes

GATA2 is the cleanest record on this beat. gene_exact:GATA2 AND organism_id:9606 AND reviewed:true returns exactly one reviewed row with primary gene name GATA2, 480 aa, and no gene synonyms in the UniProt record. Against the family: GATA1 is P15976 at 413 aa and GATA3 is P23771 at 443 aa, so the smallest gap is 37 aa and length is a usable discriminator. The one nuisance is that GATA2 P23769 and GATA3 P23771 differ by two in the accession number, which is the transcription-error pattern this project has logged twice before. The HGNC alias NFE1B carries a small slice of the older erythroid literature.

Facts and figures
uniprotP23769
gene symbolGATA2
length AA480
alphafold IDAF-P23769-F1
ensemblENSG00000179348
hgncHGNC:4171
cytoband3q21.3
protein classGATA-family dual zinc-finger transcription factor
resolution notegene_exact:GATA2 AND organism_id:9606 AND reviewed:true returned exactly ONE reviewed row, primary gene name GATA2, entry GATA2_HUMAN, 480 aa, NO gene synonyms in the UniProt record - fetched live from the UniProt REST API 2026-09-20 with gene_primary and length requested and read back. HGNC:4171 gives alias_symbol [NFE1B] and no previous symbols. Ensembl id from HGNC's single primary ensembl_gene_id, per rule 78.
paralogue discriminationGATA1 P15976 (413 aa), GATA2 P23769 (480 aa), GATA3 P23771 (443 aa) - smallest gap 37 aa, so length IS a usable discriminator here, unusually for this beat. The hazard is transcriptional rather than biological: P23769 and P23771 differ by two digits, and GATA1 and GATA3 are discussed in the same erythro-megakaryocytic and lymphoid contexts.
domains noteRead back from the UniProt flat file for P23769 on 2026-09-20 (FT ZN_FING 295..319 and 349..373). No other boundary asserted.
cell typenormal: HSCs, haematopoietic progenitors, early erythroid and megakaryocytic cells [E3]. Leukaemia: murine KMT2A-MLL3 AML of stem-cell or immature-progenitor origin [E2]; human AML blasts with subtype-dependent variability [E2]; 13 of 629 paediatric B-ALL cases as an outlier [E3].
cell statethe stemness-associated, chemoresistant state - Gata2-high leukaemia from a stem-cell origin resists doxorubicin, and the effect is p53-dampening through RASSF4 and MDM2 [E2]. In the B-ALL outliers, a myeloid-like transcriptional state inside a B-lineage blast [E3].
subcellular localizationNucleus (UniProt P23769 SUBCELLULAR LOCATION).
subnuclear localizationNOT ESTABLISHED in any leukaemia context by anything read this cycle, and not pursued further - a bounded rejection rather than a bounded absence. GATA2 chromatin occupancy is mapped by ChIP-seq in [E2] and in the heptad work [E1], but occupancy is not localisation.
named but unwritten count at dispatch16
named but unwritten count measured11
graph mention auditOf 11 node files naming GATA2 on 2026-09-20, the majority concern inv(3)/t(3;3) repositioning of the distal GATA2 enhancer to activate EVI1 with consequent GATA2 haploinsufficiency - a lesion with no KMT2A involvement - plus one note that PMID 39841459 is AML-wide and one on haemogenic endothelium. GATA2 is present in this graph as a comparator, not as a KMT2A cofactor.
what this entry does not claimThat GATA2 is a cofactor, target or dependency of a KMT2A fusion. The dependency data are the strongest negative on this beat. The defensible claim is that GATA2 marks and maintains a chemoresistant stem-like state, in AML generally, including one murine KMT2A-fusion model.
domains
[
  {
    "name": "GATA-type zinc finger 1 (N-finger)",
    "start": 295,
    "end": 319
  },
  {
    "name": "GATA-type zinc finger 2 (C-finger)",
    "start": 349,
    "end": 373
  }
]
shown for fusion
{
  "in_kmt2a_rearranged_context": "PARTIAL and weak. [E2] is a murine 'KMT2A-MLL3-driven' AML - a KMT2A-fusion context by the abstract's own description, with the construct unresolved by this beat. [E3] is a 2-patient expression correlation in KMT2A::USP2 B-ALL.",
  "wild_type_only": "[E1] normal murine haematopoiesis; [E4] inv(3)/t(3;3) AML, no KMT2A involvement.",
  "never_shown": "No GATA2 loss-of-function experiment in any KMT2A::AFF1 cell. No GATA2 measurement in infant KMT2A-rearranged material. No ChIP of a KMT2A fusion at the GATA2 locus anywhere."
}
expressed required maintains
{
  "expressed_in": "HSC/progenitor, early erythroid and megakaryocytic cells normally; subtype-variable in AML; aberrantly in ~2% of paediatric B-ALL including 2/2 KMT2A::USP2 cases.",
  "required_in": "Normal HSC generation and survival, and Runx1/Gata2 double heterozygosity is embryonic lethal [E1][E3] - a normal-tissue requirement, i.e. a toxicity statement. NOT required in KMT2A-rearranged leukaemia lines; anti-selective, positive effect in both lineage halves [E5].",
  "maintains_a_state": "The chemoresistant stem-like state via RASSF4/MDM2/p53 [E2], and a myeloid-like state in B-lineage blasts where it is aberrantly activated [E3]. This is the only register in which GATA2 has a defensible claim on this beat."
}
essentiality
{
  "source": "DepMap CRISPR (Chronos) via Open Targets GraphQL, retrieved live 2026-09-20",
  "n_all_lines": 1257,
  "chronos_mean_all_lines": 0.0311,
  "frac_lines_at_or_below_minus_0_5": 0.0199,
  "mean_other_leukaemia": -0.1897,
  "mean_kmt2a_r_myeloid": -0.0032,
  "z_kmt2a_r_myeloid": 0.49,
  "mean_kmt2a_r_b_lineage": 0.0974,
  "z_kmt2a_r_b_lineage": 0.75,
  "z_oci_aml3_npm1c": 1.36,
  "positive_control_men1_z_myeloid": -2.41,
  "genome_scan_rank": 17591,
  "genome_scan_of": 17728,
  "genome_scan_difference": 0.1997,
  "interpretation": "the most anti-selective gene on this beat and one of the most anti-selective in the genome for KMT2A-rearrangement; positive in BOTH lineage halves. Bounds fitness only - not chemoresistance, which is GATA2's published phenotype."
}
counted absence
{
  "date": "2026-09-20",
  "tool": "PubMed E-utilities esearch",
  "operand_gata2_tiab": 2447,
  "operand_mll_kmt2a_tiab": 9228,
  "intersection": 21,
  "title_checked": 20,
  "mechanistic_gata2_in_kmt2a_papers_found": 0,
  "aliases_not_searched": [
    "NFE1B"
  ],
  "bound": "titles and abstracts only; 20 of 21 ids inspected; supporting figures not searched"
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": 0.0334,
  "chronos_mean_other_leukaemia": -0.1663,
  "difference": 0.1997,
  "pan_essentiality_all_lines": 0.0307,
  "selectivity_rank": 17591,
  "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
  • Lowering GATA2 moves toward a named germline leukaemia-predisposition genotype: GATA2 haploinsufficiency causes familial AML/MDS and MonoMac and Emberger syndromes [E4]. Very few oncogenic transcription factors have a haploinsufficiency syndrome on the other side of the dose curve.
  • GATA2 is required for HSC generation and survival and Runx1/Gata2 compound heterozygosity is embryonic lethal [E1] - marrow toxicity is the mechanism, not an off-target.
  • The direction that actually helps is the reverse of a target: GATA2-high predicts chemoresistance, so the usable action is stratification plus an MDM2 inhibitor, not GATA2 inhibition [E2].
  • If the RASSF4/MDM2/p53 route is the mechanism, it is TP53-dependent by construction, and a relapsed KMT2A-rearranged leukaemia that has lost TP53 is predicted to be refractory to the whole strategy.
Still open
  1. Does GATA2 status modify chemotherapy response in KMT2A-rearranged infant B-ALL? The mechanism in [E2] is p53-dampening via RASSF4, infant t(4;11) is TP53-wild-type almost without exception, and the proposed combination - an MDM2 inhibitor with an anthracycline - is available. This is the most clinically actionable question the whole beat produced.
  2. What does a doxorubicin-conditioned CRISPR screen in KMT2A-rearranged lines show for GATA2? DepMap's unconditioned screen says GATA2 is anti-selective; the published phenotype is drug response; nobody has run the screen that would reconcile them.
  3. Is GATA2 aberrantly activated in any KMT2A::AFF1 case? The St Jude analysis found 13 outliers in 629 B-ALL and attributed two to KMT2A::USP2 - the same analysis would answer this for KMT2A::AFF1 and the cohort certainly contains such cases.
  4. Which fusion is 'KMT2A-MLL3'? [E2]'s abstract names it and this beat did not resolve the construct. If it is a KMT2A::KMT2C fusion, this graph has no entry for it, and the only GATA2-in-KMT2A functional result rests on a lesion the graph does not describe.
Every source behind this entry (6)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    GATA2 is one of the seven transcription factors in the heptad - SCL/TAL1, LYL1, LMO2, GATA2, RUNX1, ERG and FLI-1 - that combinatorially bind in blood stem/progenitor cells, and direct protein-protein interactions between RUNX1, GATA2, SCL and ERG stabilise complex binding to DNA; Runx1(+/-)::Gata2(+/-) compound heterozygous mice are not viable and show severe haematopoietic defects at midgestation.

    Combinatorial transcriptional control in blood stem/progenitor cells: genome-wide analysis of ten major transcriptional regulators · Wilson NK, Foster SD, Wang X, Knezevic K, Schütte J, Kaimakis P, Chilarska PM, Kinston S, et al. · Cell Stem Cell 7:532-544 2010
    PMID 20887958DOI
    What this evidence cannot show

    Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full; body not retrieved, so the cell system is not asserted beyond the abstract's own 'blood stem/progenitor cells'. NORMAL HAEMATOPOIESIS, murine, no fusion. Cited for heptad membership and for the Runx1/Gata2 compound-heterozygote dosage result, which is the clearest statement of GATA2's normal essentiality on this entry.

    Evidence that argues against this (1)
    • Heptad membership is a statement about normal regulatory architecture and does not transfer to leukaemia dependency. All three heptad members on this beat - ERG, GATA2 and LMO2 - are non-selective or anti-selective in the KMT2A-rearranged DepMap arm, and GATA2 is the most anti-selective of the three. analysis run by this project 2026
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In a murine model, KMT2A-MLL3-driven AML originating from a stem cell or immature progenitor population has higher Gata2 expression and is more resistant to doxorubicin; deletion of Gata2 results in a more robust induction of p53 after doxorubicin; ChIP-seq, RNA-seq and functional studies show GATA2 regulates RASSF4, a modulator of the p53 inhibitor MDM2; GATA2 and RASSF4 are anticorrelated in human cell lines and in bulk and single-cell AML patient data; Rassf4 knockdown in Gata2-low cells confers doxorubicin or nutlin-3 resistance while Rassf4 overexpression sensitises Gata2-high cells; and doxorubicin and nutlin-3 are synergistic in Gata2-high murine AML and in AML patient samples.

    GATA2 links stemness to chemotherapy resistance in acute myeloid leukemia · Alikarami F, Xie HM, Riedel SS, Goodrow HT, Barrett DR, Mahdavi L, Lenard A, Chen C, et al. · Blood 145:2179-2195 2025
    PMID 39841459DOI
    What this evidence cannot show

    Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full; body (PMC12105726) not retrieved this cycle. THE FUSION IS QUOTED AS THE ABSTRACT WRITES IT - 'KMT2A-MLL3-driven AML' - and this beat did not resolve which KMT2A::KMT2C construct that denotes or whether it is the same lesion as any KMT2A fusion this graph otherwise carries; a reader should resolve it before building on the fusion identity. MURINE for the causal arms, MYELOID throughout, with human cell lines and AML patient data used for correlation and for the synergy. No B-lineage and no infant material. The phenotype is DRUG RESPONSE, which is why the DepMap null on this gene does not contradict it - DepMap does not run under doxorubicin.

    Evidence that argues against this (1)
    • A cell-of-origin-driven expression difference that tracks drug response is, on its own, the same shape as the EVI1 cell-of-origin result on this beat, where the authors needed a knockdown to establish causality. Here the Gata2 deletion arm supplies that, but in a murine model whose fusion identity this entry could not resolve - and in human KMT2A-rearranged cell lines GATA2 knockout has a POSITIVE mean fitness effect. Broad Institute DepMap; Open Targets Platform; query run by this project 2026-09-20 2026
  3. E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    GATA2 expression is normally restricted to HSCs, haematopoietic progenitors and early erythroid and megakaryocytic cells; analysis of St Jude Cloud B-ALL transcriptomes identified aberrant GATA2 expression in 13 of 629 cases, with a significantly myeloid-like differential expression signature and BAG3 and EPOR among the targets; the correlation between KMT2A-USP2 fusion and GATA2 activation indicates a potential trans-activating mechanism and suggests GATA2 is a target of KMT2A-USP2; and GATA2 is also cis-activated, by a somatic focal deletion near GATA2 that disrupts topologically-associating-domain boundaries in one B-ALL patient.

    Aberrant GATA2 Activation in Pediatric B-Cell Acute Lymphoblastic Leukemia · Wang H, Cui B, Sun H, Zhang F, Rao J, Wang R, Zhao S, Shen S, et al. · Frontiers in Pediatrics 9:795529 2021
    PMID 35087776DOI
    What this evidence cannot show

    Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full - NOT copied from this graph's existing mech-kmt2a-usp2-gata2-trans-activation entry, per the standing rule that an identifier in the graph is unverified until this session fetched it. GRADED WEAK: the KMT2A::USP2 association rests on TWO patients, and while they are 2 of 2 in the cohort, a 2/2 rate is a striking coincidence rather than a mechanism. No ChIP-seq, no CUT&RUN and no functional experiment on the fusion. Retrospective transcriptome mining of an existing cohort. KMT2A::USP2 is a rare fusion and nothing here generalises to KMT2A::AFF1.

    Evidence that argues against this (1)
    • The same paper shows GATA2 can be activated in cis by a TAD-boundary deletion with no fusion involved, so GATA2 activation in B-ALL is not fusion-specific and at least one route to it is a structural accident. That weakens any inference that GATA2 activation implies a KMT2A-fusion mechanism. Gröschel S et al. 2014
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In AML with inv(3)/t(3;3), both 3q rearrangements reposition a distal GATA2 enhancer so that it ectopically activates EVI1 and simultaneously confers GATA2 functional haploinsufficiency - the same defect previously identified as the cause of sporadic and familial AML/MDS and of MonoMac and Emberger syndromes; genomic excision of the ectopic enhancer restored EVI1 silencing and produced growth inhibition and differentiation of AML cells, an effect replicated by pharmacological BET inhibition.

    A single oncogenic enhancer rearrangement causes concomitant EVI1 and GATA2 deregulation in leukemia · Gröschel S, Sanders MA, Hoogenboezem R, de Wit E, Bouwman BAM, Erpelinck C, van der Velden VHJ, Havermans M, et al. · Cell 157:369-381 2014
    PMID 24703711DOI
    What this evidence cannot show

    Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full; body not retrieved. THE DISEASE IS inv(3)/t(3;3) AML, NOT KMT2A-REARRANGED LEUKAEMIA - it is cited here for the direction-of-effect problem it creates for any GATA2-lowering strategy, and for the fact that it is what most of this graph's existing GATA2 mentions are about. The germline syndrome attribution (MonoMac, Emberger) is the abstract's own citation of prior work and was not independently checked this cycle.

    Evidence that argues against this (1)
    • That GATA2 haploinsufficiency predisposes to AML/MDS is a statement about germline dosage across a lifetime from conception, and this graph's standing rule is that such constraint facts are not safety signals for partial, reversible pharmacological inhibition in an adult or child. The tension with E2's chemoresistance result is therefore real but not decisive, and neither paper addresses the other. Alikarami F et al. 2025
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Fetched live on 2026-09-20, DepMap Chronos gene effect for GATA2 has mean +0.0311 across 1,257 cell lines; among leukaemia lines the KMT2A-rearranged myeloid half sits at z +0.49 and the B-lineage half at z +0.75 relative to 44 other leukaemia lines whose own mean is -0.190 - positive in both halves - and this project's genome-wide scan ranks GATA2 17,591 of 17,728 on KMT2A-selectivity with a difference of +0.1997, among the most anti-selective genes in the genome. MEN1 through the identical query shape returned z -2.41 / -2.43.

    What this evidence cannot show

    A bounded negative with a date, a route and a positive control through the identical query shape (rule 67). Seven myeloid and four B-lineage KMT2A-rearranged lines. THE BOUND: a 21-day fitness screen cannot see a CHEMORESISTANCE phenotype, and E2's entire claim is about doxorubicin response - DepMap runs no drug. It also cannot see a 2-patient subtype effect. So this bounds GATA2 as a KMT2A-selective fitness dependency and nothing else. The anti-selective direction is nonetheless informative: GATA2 is MORE required in the comparator leukaemia arm (-0.166), which is consistent with its known requirement in erythroid/megakaryocytic and stem-like lineages present in that arm. Ensembl id from HGNC primary (HGNC:4171); approvedSymbol read back as GATA2.

    Evidence that argues against this (1)
    • Published GATA2 work in leukaemia is about drug response and about state, not about fitness, so a fitness screen is close to the wrong instrument. The result that WOULD be decisive - DepMap's drug-response arm, or a doxorubicin-conditioned CRISPR screen in KMT2A-rearranged lines - has not been run by anyone, including this project. Alikarami F et al. 2025
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    Counted absence, measured 2026-09-20 with PubMed E-utilities esearch: GATA2[tiab] returns 2,447 records and (MLL OR KMT2A)[tiab] returns 9,228, while GATA2[tiab] AND (MLL OR KMT2A)[tiab] returns 21. Twenty of the 21 ids were title-checked by efetch and exactly TWO are on topic - PMID 39841459 (GATA2 links stemness to chemotherapy resistance in AML, this entry's E2) and PMID 35087776 (Aberrant GATA2 activation in pediatric B-ALL, this entry's E3). The other eighteen are mutational-landscape and CEBPA-subgroup cohort papers, an IgA-nephropathy bioinformatics paper, a mesothelin immunohistochemistry paper, a case report of a GATA2-MECOM AML that later acquired a KMT2A-ASXL2 fusion, an Eed/MLL-AF9 paper and an Arid1b/Arid2 MLL-AF9 paper. No GATA2 loss-of-function experiment in a KMT2A-rearranged cell appears anywhere in the set.

    What this evidence cannot show

    Bounded to titles and abstracts; supporting figures and supplementary tables not searched (rule 80). GRADED MODERATE rather than established because 20 of the 21 ids were inspected, not all 21 - the esearch call returned 20 ids at the retmax used - so this is a near-complete rather than a complete enumeration and is reported as such. Aliases: NFE1B was NOT searched against the KMT2A clause. Both operand counts are given so the small intersection has denominators. A CORRECTION MADE WHILE WRITING: the first draft of this item asserted that PMID 35087776 was absent from the returned set; it is present, and the assertion was made from a misreading of the id list rather than from the list itself.

    Evidence that argues against this (1)
    • This absence is about mechanistic studies, not about the gene's relevance. GATA2 appears in 11 node files of this graph and the majority concern inv(3)/t(3;3) GATA2-enhancer hijacking of EVI1 - a real and important leukaemia mechanism that simply is not a KMT2A mechanism. The correct reading is a misfiled gene, not an unstudied one. Gröschel S et al. 2014
Connected to