No menin interaction outside the chromatin complex has ever been tested in a haematopoietic cell - 0 of 9, and the compartment is the wrong one in every case
also called Q128 bounded absence · menin non-chromatin partners never tested in blood
Every experiment behind menin's nine non-chromatin partners was done in kidney, colon, bone, lung or skin cancer cells. Not one was done in a blood cell, which is the only place this project's disease exists.
The same thing, for a scientist
A complete enumeration of the host systems behind all 22 IntAct records covering the nine partners returns **five cell lines and zero haematopoietic cells**: HEK293T embryonic kidney (18 records), HCT116 colon carcinoma (1), U2OS osteosarcoma (1), A375 melanoma (1), H1299 lung carcinoma (1, written 'H199' by IntAct) [E1]. **There is no primary cell, no patient sample, no leukaemia line and no mouse in the entire evidence base.** Every record is a tagged over-expression system, and only one - the Mo Figure 4D semi-IP - has an endogenous protein on either side [E1]. Separately, **no record of any kind involves a menin inhibitor**: not one of the 22 rows was measured under revumenib, ziftomenib or any tool compound, so nothing here says whether a menin inhibitor would disturb any of these associations [E2]. The absence is classed `never-measured` and the class is the point: no archive holds this, and a literature query cannot close it.
The complete host enumeration, with nothing dropped
| host system | records | partners |
|---|---|---|
| HEK293T embryonic kidney | 18 | AKT1, EGFR, PTEN, CD226, BRAF x2 (BRET, GST pull-down), FBXW7 x4, SMAD4 x4, SPOP x4 |
| HCT116 colon carcinoma | 1 | ZC3HC1 |
| U2OS osteosarcoma | 1 | CD226 |
| A375 malignant melanoma | 1 | BRAF (NanoPCA) |
| H1299 lung carcinoma (IntAct writes 'H199') | 1 | BRAF (semi-IP, endogenous menin) |
| haematopoietic, any | 0 | none |
22 records, 5 lines, 0 blood (18 + 1 + 1 + 1 + 1 = 22). Rule 78 applies and is honoured: nothing was dropped from this join. Every one of the 22 rows returned by the IntAct REST filter is accounted for in the table above; there is no residual and no unresolved host string.
Why this is worth its own entry rather than a caveat on nine others. The project's standing concern is evidence imported from a non-index system, and the usual case is a mouse model or a different fusion. This is more extreme: the index system is not merely under-represented, it is completely absent, across two decades of two large platforms, for the entire non-chromatin half of menin's curated interactome.
And no menin inhibitor anywhere in the evidence base
None of the 22 records was measured in the presence of a menin inhibitor. That is unsurprising - an interactome screen has no reason to add one - but it has a consequence the project should state rather than assume.
The menin-KMT2A interface is a pocket, and revumenib and ziftomenib occupy it. If any of these nine associations is real and uses the same surface, the inhibitor would disrupt it and that disruption would be an unmeasured off-target-within-target effect. If an association uses a different surface, the inhibitor would leave it alone, and a patient's non-chromatin menin biology would continue unchanged under therapy. No record maps a region, residue or domain on either protein for any of the nine, so which of those two states obtains is not merely unmeasured - it is unposable from the current data.
This is the concrete sense in which the beat's founding question - if menin sits in these complexes, menin's role is not only chromatin, and that bears on what a menin inhibitor does - is unanswered rather than answered negatively.
What would close it, and why a literature query cannot
never-measured is the expensive class and this entry does not pretend otherwise. The archives were considered and are the wrong instrument here: an interaction in a primary blast is not deposited anywhere, because no one has generated it. A GEO or PRIDE query would return the existing AP-MS and proximity datasets, all in the same non-haematopoietic lines.
The one archive question that IS worth asking, and which this beat did not ask: does any deposited menin-bait AP-MS or BioID dataset exist in a leukaemia line? This corpus already records that the index fusion's own AP-MS interactome (PMID 29777171) is measured-not-curated - published and absent from IntAct. If a menin-bait experiment in a leukaemia line exists in the same uncurated state, the whole framing of this entry changes, and finding out is a PRIDE query rather than a bench project. That is named here as the next step precisely because this beat did not take it, and an absence entry that does not say which archive query it skipped is the weaker kind this project counted 24 of.
| script | no committed script; IntAct REST host enumeration, 2026-09-23 |
|---|---|
| status | run |
| result summary | 22 records, 5 cell lines, 0 haematopoietic systems, 0 primary cells, 0 animals, 1 endogenous side, 0 menin-inhibitor conditions, 0 mapped interfaces. |
| nothing dropped | true |
| n records | 22 |
| records with an endogenous side | 1 |
| records under a menin inhibitor | 0 |
| records with a mapped interface | 0 |
| absence class | never-measured |
| absence class note | No experiment has tested any of menin's nine non-chromatin IntAct partners in a haematopoietic cell. This cannot be closed from an archive because the data were never generated - every deposited dataset behind these rows is in the same five non-haematopoietic lines. THE ONE ARCHIVE QUERY THIS BEAT DID NOT RUN, and which should be run before any bench work: does a deposited menin-bait AP-MS or BioID dataset exist in a leukaemia line, in the same uncurated state as PMID 29777171? That is a PRIDE / ProteomeXchange question and it is named here because skipping it is the defect this project counted in 24 of 26 absence entries. |
inputs
[ "https://www.ebi.ac.uk/intact/ws/interaction/findInteractions/O00255?page=0&pageSize=1000" ]
host enumeration — the numbers behind the chart
{
"HEK293T embryonic kidney": 18,
"HCT116 colon carcinoma": 1,
"U2OS osteosarcoma": 1,
"A375 malignant melanoma": 1,
"H1299 lung carcinoma (IntAct: 'H199')": 1,
"haematopoietic": 0
}- The whole clinical relevance of a non-chromatin menin function is that a menin inhibitor might disturb it, and zero of the 22 records were measured under an inhibitor while zero map an interface. So the drug question is not merely unanswered - the data cannot be made to bear on it without new experiments.
- The reverse risk is that this absence is quoted as reassurance. 'Nobody has shown menin does anything outside chromatin in a blood cell' is not 'menin does nothing outside chromatin in a blood cell', and an entry this clean is exactly the kind that gets compressed into the second sentence.
- Does a deposited menin-bait AP-MS or BioID dataset exist in a leukaemia line, curated or not? Naming this as unrun is the honest state of this entry; running it is a PRIDE query and would either close or transform the absence.
- Is there a menin proximity-labelling dataset in ANY primary human cell? Proximity labelling is the method most likely to have been applied to a low-abundance nuclear protein in a hard-to-transfect cell, and the corpus has not looked.
- If menin's non-chromatin interactome were mapped properly in a blast, would the nine reappear? A single BioID in a KMT2A-rearranged line would test all nine at once and is the highest-value single experiment this beat can name.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
All 22 IntAct records covering the nine partners, retrieved live from the REST endpoint on 2026-09-23, carry hostOrganism values resolving to five cell lines: HEK293T embryonic kidney (18 records), HCT116 colon carcinoma (1), U2OS osteosarcoma (1), A375 malignant melanoma (1) and H1299 lung carcinoma (1, recorded by IntAct as 'H199 lung lymph node carcinoma cell'). None is a haematopoietic cell, primary cell, patient sample or animal. Only one record, the Mo et al. Figure 4D semi-immunoprecipitation, has an endogenous untagged protein on either side.
IntAct molecular interaction database, REST findInteractions on O00255, retrieved 2026-09-23 by Q128 · IntAct (EMBL-EBI) 2026What this evidence cannot show
This bounds IntAct, not the literature: an interaction published and never curated is invisible to it, and this corpus already holds one such case (PMID 29777171). The enumeration is complete with respect to the 22 rows returned and nothing was dropped, but the 22 rows are themselves the output of a partner filter over a single database at a single moment.
Evidence that argues against this (1)
- HEK293T is the standard chassis for interaction proteomics precisely because it is transfectable and well-characterised, and a great many real complexes were first found there. 'Not a blood cell' is a relevance objection, not a validity objection, and conflating the two would let this project discard most of the structured interactome. Huttlin EL et al. 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
No record among the 22 was measured in the presence of a menin inhibitor, and no record maps an interface region, residue or domain on either menin or the partner. IntAct's featureRanges and featureTypes fields for these rows carry only epitope tags and, on the Mo rows, partner-side point mutations.
IntAct molecular interaction database, REST findInteractions on O00255, retrieved 2026-09-23 by Q128 · IntAct (EMBL-EBI) 2026What this evidence cannot show
Absence of a drug condition in a curated interaction record is close to uninformative on its own - interactome screens do not include drugs - so the load-bearing half of this claim is the unmapped interface, which is read directly from the featureRanges field. Whether any UNCURATED experiment has combined a menin inhibitor with any of these partners was not searched by this beat.
Evidence that argues against this (1)
- The menin inhibitor point would be settled the other way by any published study showing a menin-inhibitor-sensitive non-chromatin menin function, and this beat ran no literature query for one. The claim here is strictly about the 22 IntAct records, and reading it as 'no such study exists' would exceed the evidence. Mo X et al. 2022
the system-level bound under the per-partner adjudications
zero of menin's nine non-chromatin IntAct partners has been tested in a haematopoietic cell
none of the 22 records was measured under a menin inhibitor and none maps an interface, so the drug question is unposable from these data
same bound: no record under any menin inhibitor