HOXA7 - the one HOXA paralogue besides HOXA9 with a published loss-of-function requirement for MLL-fusion transformation, and the requirement was shown in vitro only
also called HOX1A · HOX1 · Homeobox protein Hox-A7 · Homeobox protein Hox-1.1 · Homeobox protein Hox-1A
HOXA7 is a gene sitting right next to HOXA9 in the block of genes that KMT2A-fusion leukaemia switches on, and in mouse cells an MLL fusion cannot immortalise blood progenitors without it. That single experiment is the whole case: everywhere else HOXA7 is simply measured alongside HOXA9 and moves with it, which is what neighbours in this gene block always do.
The same thing, for a scientist
HOXA7 (P31268, 230 aa, 7p15.2) is an anterior-group HOXA paralogue transcribed from the same collinear, shared-enhancer, shared-TAD cluster as HOXA9 and HOXA10. It is named in 17 entries of this graph and had no entry of its own. **The one thing that distinguishes it from a passenger is Ayton and Cleary 2003**: in primary murine myeloid progenitors, Hoxa7 and Hoxa9 were each specifically required for efficient in vitro immortalisation by an MLL fusion protein, and not by other leukaemogenic fusions [E1]. The in vivo arm of that same paper tested **HOXA9 only**, so HOXA7's requirement is bounded to a colony/immortalisation assay and has never been tested in a transplanted animal, in a human cell, or in a B-lineage or KMT2A::AFF1 context [E1]. Independently, proviral tagging in BXH-2 mice activates *Hoxa7* and *Hoxa9* as distinct common integration sites, each correlated with *Meis1* activation - genetic selection for HOXA7 specifically, not co-expression [E2]. In human t(4;11) cells, HOXA7 falls with HOXA9 and MEIS1 when MLL-AF4 is knocked down [E3], and it is *suppressed* rather than activated by at least one KMT2A fusion, KMT2A::SEPTIN6 [E4] - so 'the fusion turns on HOXA7' is not true of every fusion. HOXA7 is **not a dependency**: in this project's own genome-wide DepMap scan it ranks 1,104 of 17,728 on KMT2A-selectivity with a difference of only -0.102, and a live re-query gives mean Chronos -0.0616 across 1,255 lines with 0.56% of lines below -0.5 [E5]. The lineage split is the informative part: z -0.96 in the KMT2A-r myeloid half, -0.53 in the B-lineage half, and **-2.88 in OCI-AML3**, an NPM1c, KMT2A-wild-type, HOXA-high line - HOXA7 dependency tracks the HOXA-high STATE, not the fusion [E5].
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.
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.
Expressed in / required in / maintains a state - kept separate on purpose
This project carries analyses/lineage-confound-myeloid-tf-dependencies/ because these three were pooled once and the pooling produced two false dependencies. For HOXA7:
- Expressed in. Human t(4;11) B-ALL lines SEM and RS4;11 express HOXA7 at a level that falls measurably on MLL-AF4 knockdown 3; murine myeloid progenitors immortalised by MLL fusions carry a characteristic Hoxa cluster profile that includes Hoxa7 1. Expression is well documented and is the weakest of the three statements, because the HOXA cluster is collinear and shares enhancers and a TAD - see
concept-hoxa-expression-is-collinear-with-reciprocal-statusandgenome-ctcf-cohesin-hoxa-tad-boundary. Co-expression with HOXA9 is the null expectation here, not evidence. - Required in. One system only: efficient in vitro myeloid immortalisation of primary murine progenitors by an MLL fusion protein 1. Not required-in-vivo (untested for HOXA7), not required in human cells (untested), not required in B-lineage (untested), not a CRISPR fitness dependency in any of 1,255 human cancer lines 5.
- Maintains a state. The state in question is the myeloid clonogenic-progenitor Hoxa profile that MLL fusions enforce and whose continued maintenance is associated with continued immortalisation 1. HOXA7 is part of that profile. Whether it maintains the state or merely reports it is exactly what the missing in vivo and human experiments would decide.
Cluster member or independent driver - what separates the two, and what the evidence actually reaches
Collinear activation of the HOXA cluster means HOXA7 is expected to move with HOXA9 whether or not it does anything. Two lines of evidence go beyond co-expression, and both are murine and myeloid:
- Genetic selection. BXH-2 retroviral insertional mutagenesis identifies Hoxa7 as a common viral integration site in its own right, alongside Hoxa9 and Meis1, with Hoxa7 or Hoxa9 activation strongly correlated with Meis1 activation 2. A provirus landing in Hoxa7 and being selected for is a statement about Hoxa7, not about the cluster.
- Loss of function. Hoxa7 was specifically required for efficient MLL-fusion immortalisation, and the requirement was absent for other leukaemogenic fusions tested in the same work 1.
What neither reaches. A HOXA7-selective loss-of-function in a human KMT2A-rearranged cell, of any lineage. A HOXA7 knockout in a transplant model. Any experiment separating HOXA7 from HOXA9 in B-lineage KMT2A::AFF1 - the index disease. And nothing at all on whether HOXA7 loss can be compensated by HOXA9, which is the obvious buffer and which E5's near-null CRISPR score cannot distinguish from dispensability.
The dependency number, split by lineage, and the line that makes the point
Re-queried live on 2026-09-20 through the same Open Targets/DepMap route this project uses elsewhere, with MEN1 carried through the identical query shape as a positive control (MEN1 returned z -2.41 myeloid / -2.43 B-lineage, reproducing analyses/lineage-confound-myeloid-tf-dependencies/ exactly, so the instrument was working) 5:
| arm | n lines | mean Chronos | z vs other-leukaemia |
|---|---|---|---|
| KMT2A-r myeloid | 7 | -0.195 | -0.96 |
| KMT2A-r B-lineage | 4 | -0.136 | -0.53 |
| other leukaemia | 44 | -0.063 | 0 by construction |
| all lines | 1,255 | -0.062 | - |
| OCI-AML3 (NPM1c, KMT2A-WT, HOXA-high) | 1 | - | -2.88 |
| U-937 (PICALM::MLLT10, HOXA-high) | 1 | - | -0.17 |
HOXA7 is the deepest of the six factors on this beat and it is still shallow. The striking row is OCI-AML3: the single strongest HOXA7 effect in the panel is in a line with no KMT2A rearrangement at all. Whatever HOXA7 is doing, the CRISPR data associate it with the HOXA-high state (mechanism-hoxa-high-vs-hoxa-low) rather than with the fusion. Two caveats bound this hard: one line is one line, and paralogue redundancy makes a single-gene HOXA7 knockout uninterpretable as evidence of dispensability by construction.
One KMT2A fusion SUPPRESSES HOXA7 - so 'the fusion drives HOXA7' is a statement about some fusions
KMT2A::SEPTIN6 in the KOPM-88 AML line activates BMP signalling and inhibits expression of HOXA7 and HOXA9 4. That is the opposite sign to every other KMT2A fusion this graph records, and it is the only published transcriptional characterisation of that fusion. It does not overturn the HOXA7 story; it bounds it to the fusions where the sign has been measured, which is a small set.
Identifier hazards for HOXA7 specifically
- HOXA7 (P31268, 230 aa) and HOXB7 (P09629, 217 aa) are 13 residues apart - inside rule 66's SMC1A/SMC3 limit of 16, so
verify-accessions.pycannot reliably catch a paralogue swap between the two clusters. Both are anterior-group paralogues and both are discussed in HOX-driven leukaemia. - HOXA9 is P31269 and HOXA7 is P31268: the accessions differ in the final digit, exactly the pattern that already cost this project a HOXA9/HOXA10 note (P31269 vs P31260).
- Retired official symbols HOX1A and HOX1 carry the pre-1990s literature;
HOX1alone is unusable as a search term. - A retraction sits in this literature. PMID 21925392 (Mol Cell 2011, the lncRNA Mistral activating Hoxa6/Hoxa7 by recruiting MLL1) is flagged by PubMed as a Retracted Publication. It is the single most on-topic-looking HOXA7-and-MLL1 title in the whole search and it is not citable. It is not currently cited anywhere in this graph; recorded here so the next reader does not add it.
| uniprot | P31268 |
|---|---|
| gene symbol | HOXA7 |
| length AA | 230 |
| alphafold ID | AF-P31268-F1 |
| ensembl | ENSG00000122592 |
| hgnc | HGNC:5108 |
| cytoband | 7p15.2 |
| protein class | anterior-group HOX homeodomain transcription factor |
| resolution note | gene_exact:HOXA7 AND organism_id:9606 AND reviewed:true returned exactly ONE reviewed row, primary gene name HOXA7, entry HXA7_HUMAN, 230 aa, synonym HOX1A - fetched live from the UniProt REST API 2026-09-20 with gene_primary and length requested and read back. Ensembl id taken from HGNC's single primary ensembl_gene_id (rule 78), not from a live multi-mapping resolver. |
| paralogue discrimination | HOXA7 P31268 (230 aa) against HOXB7 P09629 (217 aa) is a 13 aa gap - INSIDE rule 66's SMC1A/SMC3 limit of 16, so the length check in verify-accessions.py cannot be relied on to catch a cross-cluster swap. HOXA9 is P31269, one digit from P31268. Assert gene_primary. |
| domains note | Homeodomain boundaries read back from the UniProt flat file for P31268 on 2026-09-20 (FT DNA_BIND 130..189). No other boundary is asserted; the hexapeptide/PBX-interaction motif was not read at residue level this cycle and is therefore omitted rather than guessed. |
| cell type | primary murine myeloid progenitors (the immortalisation requirement, E1); human t(4;11) B-ALL lines SEM and RS4;11 (expression downstream of the fusion, E3); murine BXH-2 myeloid leukaemia (proviral selection, E2) |
| cell state | the MLL-fusion-enforced myeloid clonogenic-progenitor Hoxa profile - the differentiation-blocked, self-renewing state (E1). No published measurement in a quiescent or a differentiating state. |
| subcellular localization | Nucleus (UniProt P31268 SUBCELLULAR LOCATION, no experimental PubMed evidence code attached to the statement in the record read 2026-09-20). |
| subnuclear localization | UNPUBLISHED as far as this cycle could establish. No imaging of HOXA7 subnuclear distribution - speckles, polycomb bodies, nucleolus, lamina-associated domains - was found in any system, let alone a KMT2A-rearranged one. This is a bounded absence on an entity rather than a measurement: HOXA7[tiab] returns 249 records and none of the titles surfaced in this beat's searches concerns subnuclear distribution; supporting figures were not searched. |
| named but unwritten count at dispatch | 50 |
| named but unwritten count measured | 17 |
| count discrepancy note | The dispatch brief gave 50 entries mentioning HOXA7. A word-boundary grep over data/graph/nodes on 2026-09-20 returns 17 node files. The larger number presumably counts docs/, analyses/ and substring matches. The 17 node mentions were read and are genuinely on-beat. |
| what this entry does not claim | That HOXA7 is required for KMT2A::AFF1 leukaemia, or for any leukaemia in vivo, or in any human cell. The published requirement is one in vitro immortalisation assay in murine myeloid progenitors. |
domains
[
{
"name": "Homeobox (homeodomain)",
"start": 130,
"end": 189
}
]shown for fusion
{
"in_kmt2a_rearranged_context": "PARTIAL. The loss-of-function requirement (E1) is in murine primary myeloid progenitors immortalised by an MLL fusion protein - a KMT2A-fusion context, but murine, myeloid, in vitro, and not KMT2A::AFF1. The expression dependence on the fusion (E3) is human, B-lineage and KMT2A::AFF1, but is co-regulation only.",
"wild_type_only": "E2 (BXH-2 proviral tagging) is a retroviral model with no MLL fusion at all.",
"never_shown": "No HOXA7 loss-of-function in a human KMT2A-rearranged cell of any lineage. No HOXA7 in vivo requirement in any MLL-fusion transplant model. No HOXA7 experiment in infant KMT2A::AFF1 material."
}expressed required maintains
{
"expressed_in": "Human t(4;11) B-ALL (SEM, RS4;11) and murine MLL-fusion-immortalised myeloid progenitors; expression falls when the fusion is knocked down [E3] and is enforced by the fusion [E1]. Collinear cluster co-expression with HOXA9/HOXA10 is the null expectation.",
"required_in": "Efficient in vitro myeloid immortalisation of primary murine progenitors by an MLL fusion protein [E1]. Nowhere else. Not a CRISPR fitness dependency in 1,255 human lines [E5].",
"maintains_a_state": "Part of the MLL-enforced myeloid clonogenic-progenitor Hoxa profile whose maintenance accompanies continued immortalisation [E1]. Whether HOXA7 maintains that state or reports it is undecided."
}essentiality
{
"source": "DepMap CRISPR (Chronos) via Open Targets GraphQL, retrieved live 2026-09-20",
"n_all_lines": 1255,
"chronos_mean_all_lines": -0.0616,
"frac_lines_at_or_below_minus_0_5": 0.0056,
"n_other_leukaemia_lines": 44,
"mean_other_leukaemia": -0.0627,
"n_kmt2a_r_myeloid": 7,
"mean_kmt2a_r_myeloid": -0.1954,
"z_kmt2a_r_myeloid": -0.96,
"n_kmt2a_r_b_lineage": 4,
"mean_kmt2a_r_b_lineage": -0.1355,
"z_kmt2a_r_b_lineage": -0.53,
"z_oci_aml3_npm1c": -2.88,
"z_u937_picalm_mllt10": -0.17,
"positive_control_men1_z_myeloid": -2.41,
"positive_control_men1_z_b_lineage": -2.43,
"genome_scan_rank": 1104,
"genome_scan_of": 17728,
"genome_scan_difference": -0.1016,
"interpretation": "not a dependency; top-decile KMT2A-selectivity on a trivial effect size; the deepest single effect in the panel is in a KMT2A-wild-type HOXA-high line"
}retraction hazard
{
"pmid": "21925392",
"status": "Retracted Publication per PubMed PublicationType, checked 2026-09-20",
"why_it_matters": "It is titled 'The noncoding RNA Mistral activates Hoxa6 and Hoxa7 expression and stem cell differentiation by recruiting MLL1 to chromatin' - the single most on-topic-looking HOXA7-and-MLL1 record in the literature. Not cited anywhere in this graph as of 2026-09-20; logged so it is not added."
}depmap kmt2a selectivity
{
"chronos_mean_kmt2a_rearranged": -0.1736,
"chronos_mean_other_leukaemia": -0.072,
"difference": -0.1016,
"pan_essentiality_all_lines": -0.0613,
"selectivity_rank": 1104,
"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."
}- The only chemical matter that reaches HOXA7 is the shared HOX-PBX hexapeptide interface, which is paralogue-blind by design - an agent hitting it cannot distinguish HOXA7 from HOXA9, HOXA10, or HOX proteins running normal gut, limb and marrow development.
- HOXA7 has no dependency signal in any human line [E5], so there is no potency read-out against which a HOXA7-directed agent could be optimised. This is the same drug-development dead end PBX3 and HOXA10 have in this graph.
- If HOXA7 and HOXA9 are mutually redundant, a HOXA9-selective strategy has a resistance route 30 kb away inside the same TAD and under the same enhancers - and nobody has looked for a paralogue swap under menin inhibition.
- KMT2A::SEPTIN6 SUPPRESSES HOXA7 [E4]. Any HOXA-directed therapeutic logic has to say which fusions it applies to, because at least one runs the other way.
- This entry's strongest evidence is a single 2003 murine in vitro paper. A strategy resting on it is resting on one experiment nobody has repeated.
- Does Hoxa7 loss block MLL-fusion leukaemogenesis IN VIVO? The founding paper ran that arm for Hoxa9 and not for Hoxa7, and twenty-three years later nobody has run it. It is a single transplant experiment and it would convert HOXA7 from an in vitro requirement into a driver or into a passenger.
- Is HOXA7 required in a B-lineage KMT2A::AFF1 cell? Every loss-of-function datum on this gene is murine and myeloid; the index disease has never been tested.
- Can HOXA9 rescue HOXA7 loss, and vice versa? A combinatorial HOXA7/HOXA9 knockout in one KMT2A-rearranged cell would decide whether the near-null CRISPR score means dispensable or buffered - and it is the experiment that the whole posterior/anterior HOXA question turns on.
- Why is the strongest HOXA7 CRISPR effect in the panel in OCI-AML3, an NPM1c KMT2A-wild-type line? If HOXA7 dependency tracks the HOXA-high state rather than the fusion, that reframes it as a state dependency shared across NPM1c, NUP98-fusion, CALM-AF10 and KMT2A-rearranged disease - and predicts a menin inhibitor would collapse it in all of them.
- What does HOXA7 occupancy look like? There is no published HOXA7 ChIP-seq or CUT&RUN in any leukaemia this beat could find, so the graph asserts a HOXA programme without knowing whether HOXA7 and HOXA9 bind the same sites.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In primary murine myeloid progenitors, MLL oncoproteins enforce a characteristic Hoxa cluster expression profile whose continued maintenance is associated with conditional MLL-associated immortalisation, and Hoxa7 and Hoxa9 were specifically required for efficient in vitro myeloid immortalisation by an MLL fusion protein but not by other leukaemogenic fusion proteins; in a bone-marrow transduction/transplantation model, HOXA9 - and only HOXA9 - was tested and found essential for MLL-dependent leukaemogenesis in vivo.
Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9 · Ayton PM, Cleary ML · Genes & Development 17:2298-2307 2003What this evidence cannot show
Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full; the body was not retrieved this cycle, so the exact knockout allele, the fusion used for each arm and the effect sizes are not asserted here. MURINE, MYELOID, PRIMARY PROGENITORS, IN VITRO for the HOXA7 arm - rule 73 applies and is written into the claim. The in vivo requirement stated in the same abstract is for HOXA9 alone; extending it to HOXA7 would be exactly the generalisation this graph's rule 73 exists to stop. Nothing here is B-lineage and nothing here is KMT2A::AFF1.
Evidence that argues against this (1)
- A CRISPR knockout of HOXA7 has no fitness effect in any of 1,255 human cancer cell lines including 11 KMT2A-rearranged leukaemia lines, so whatever requirement exists in primary murine progenitors does not reproduce as a dependency in established human lines - either because the assay differs (immortalisation versus 21-day fitness), or because HOXA9 buffers the loss, or because the requirement is species- or system-specific. Broad Institute DepMap; Open Targets Platform; query run by this project 2026-09-20 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In murine BXH-2 myeloid leukaemias, proviral tagging identified Hoxa7, Hoxa9 and the Pbx1-related homeobox gene Meis1 as genes whose expression is activated by proviral integration, and proviral activation of Hoxa7 or Hoxa9 is strongly correlated with proviral activation of Meis1, implying cooperation.
Cooperative activation of Hoxa and Pbx1-related genes in murine myeloid leukaemias · Nakamura T, Largaespada DA, Shaughnessy JD, Jenkins NA, Copeland NG · Nature Genetics 12:149-153 1996What this evidence cannot show
Abstract fetched from PubMed on 2026-09-20; full text not retrieved and the number of Hoxa7 integrations is therefore NOT asserted here. MURINE, MYELOID, retrovirus-induced - not an MLL-fusion system at all. It is cited for one thing only: that Hoxa7 is selected for as an integration site in its own right, which is evidence against HOXA7 being purely a collinear passenger. It says nothing about human disease, about KMT2A rearrangement, or about B-lineage.
Evidence that argues against this (1)
- Proviral selection for Hoxa7 in a retroviral model says a Hoxa7 gain-of-function can contribute to murine myeloid leukaemia; it does not say HOXA7 is required in a leukaemia driven by something else. The only loss-of-function evidence is E1's in vitro immortalisation arm, and the in vivo arm of that same paper tested HOXA9 rather than HOXA7. Ayton PM et al. 2003
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In the human t(4;11) KMT2A::AFF1 B-ALL line RS4;11, ectopic miR-142-3p reduces MLL-AF4 expression, suppresses proliferation, induces apoptosis, and strongly reduces the expression of the MLL-AF4 target genes HOXA9, HOXA7 and HOXA10 - so HOXA7 transcription in a human cell carrying the index fusion moves with the fusion.
MicroRNA-142-3p inhibits cell proliferation in human acute lymphoblastic leukemia by targeting the MLL-AF4 oncogene · Dou L, Li J, Zheng D, Li Y, Gao X, Xu C, Gao L, Wang L, et al. · Molecular Biology Reports 40:6811-6819 2013What this evidence cannot show
Abstract fetched from PubMed efetch XML on 2026-09-20 and read in full; figures not examined and the magnitude of the HOXA7 reduction is NOT asserted. CORRECTION MADE WHILE WRITING: this beat first drafted the claim as 'SEM and RS4;11' from this graph's own prose, and the fetched abstract names RS4;11 only - an instance of the standing rule that a citation re-used from a sibling entry is unverified until re-fetched. The reagent is a microRNA mimic, which is inherently multi-target: a miR-142-3p effect on HOXA7 is not necessarily routed through MLL-AF4. HUMAN, B-LINEAGE, the index fusion - which is why it is here - but it establishes co-regulation and not requirement, and a single cell line.
Evidence that argues against this (1)
- A microRNA is a pleiotropic reagent and a fall in HOXA7 after miR-142-3p transfection is consistent with any number of routes that do not pass through MLL-AF4. The cleaner statement in the same direction is E1's, where the Hoxa profile is enforced by a conditional MLL fusion - and that is murine and myeloid. Ayton PM et al. 2003
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In the paediatric AML cell line KOPM-88, the fusion gene KMT2A::SEPTIN6 activates BMP signalling and INHIBITS expression of HOXA7 and HOXA9 - the opposite sign from every other KMT2A fusion for which HOXA output has been measured.
Fusion Gene KMT2A::SEPTIN6 in Acute Myeloid Leukemia Cell Line KOPM-88 · Nagel S, Meyer C, Kaufmann M, Fähnrich S, Rand U, Pommerenke C, MacLeod RAF, Eberth S · Cells 15(14):1286 2026What this evidence cannot show
Abstract fetched from PubMed on 2026-09-20 and read in full; figures not examined. ONE cell line, KOPM-88, the only published model of this fusion, derived from a boy at relapse - so a line-specific effect cannot be excluded and there is no second model to check against. The direction (suppression) is stated in the abstract; the magnitude is not asserted here. This is MYELOID and a rare fusion, not KMT2A::AFF1.
Evidence that argues against this (1)
- A single cell line carrying a rare fusion, characterised in one paper, is thin ground on which to generalise about fusion heterogeneity; the KOPM-88 result may be a property of that line's other lesions rather than of KMT2A::SEPTIN6. It is recorded because a counter-example of the right shape is worth more than its weight, not because it is strong. Ayton PM et al. 2003
- 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 HOXA7 has mean -0.0616 across 1,255 cell lines with 0.56% of lines at or below -0.5; among leukaemia lines the KMT2A-rearranged myeloid half sits at z -0.96 and the B-lineage half at z -0.53 relative to 44 other leukaemia lines, while OCI-AML3 - an NPM1c, KMT2A-wild-type, HOXA-high line - sits at z -2.88. MEN1 carried through the identical query shape returned z -2.41 myeloid and -2.43 B-lineage, reproducing this project's existing lineage-split analysis.
DepMap CRISPR (Chronos) gene effect for HOXA7 (ENSG00000122592) and MEN1 (ENSG00000133895), retrieved via the Open Targets Platform GraphQL API, with the lineage split of analyses/lineage-confound-myeloid-tf-dependencies re-run for this gene · Broad Institute DepMap; Open Targets Platform; query run by this project 2026-09-20 · Database query 2026What this evidence cannot show
A bounded negative with a date, a query route and a positive control (MEN1 through the identical query shape, per rule 67). Seven myeloid and four B-lineage KMT2A-rearranged lines, so one line moves a mean; OCI-AML3 is a single line and its z of -2.88 is a one-line observation, not a result. A database fact opposable on scope: immortalised adult-derived lines, none infant, roughly 21 days, no self-renewal or differentiation read-out, and paralogue redundancy in the HOXA cluster makes a single-gene HOXA7 knockout uninterpretable as evidence of dispensability. Ensembl id resolved through HGNC (HGNC:5108 primary ensembl_gene_id) per rule 78, and the Open Targets record's approvedSymbol was read back as HOXA7 before any number was used.
Evidence that argues against this (1)
- This project's own genome-wide scan over 17,728 genes ranks HOXA7 1,104th on KMT2A-selectivity with a difference of -0.1016 (KMT2A-r mean -0.1736 against -0.072 in other leukaemia lines) - top decile but a tiny effect, and the same scan puts MEN1 first at -0.8656, eight-fold larger. HOXA7 is selective-ish and trivial in size, and a z-based ranking in this tail is close to a ranking of narrowness rather than of effect. analysis run by this project 2026
top-decile baseline, retains 1.06% absolutely at d9
- regulates →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
HOXA7 transcription falls when MLL-AF4 is knocked down in RS4;11 (co-regulation, not requirement)
- co-occurs with →HOXA9 — required for survival of established MLL-rearranged leukemia, and dispensable for starting it
adjacent HOXA paralogues, collinearly co-expressed and jointly required for in vitro MLL-fusion immortalisation
both reduced together with HOXA9 when MLL-AF4 is depleted
- co-occurs with →MEIS1 — rate-limiting for MLL leukemia stem cells, a surviving co-dependency, and embryonic-lethal when removed
proviral activation of Hoxa7 is strongly correlated with proviral activation of Meis1 in BXH-2 murine myeloid leukaemia
member of the MLL-fusion-enforced myeloid clonogenic-progenitor Hoxa profile
KMT2A::SEPTIN6 in KOPM-88 suppresses HOXA7 and HOXA9 - opposite sign to other KMT2A fusions
- related to →HOXA-high versus HOXA-low KMT2A-rearranged leukemia — the split that breaks the one-mechanism story
the deepest HOXA7 CRISPR effect in the panel is in OCI-AML3, a KMT2A-wild-type HOXA-high line
collinear cluster activation is the null expectation against which any HOXA7-specific claim must be read
HOXA7 ranks 1,104 of 17,728 on KMT2A-selectivity, difference -0.1016