IRX1 occupancy under PARTIAL menin inhibition in KMT2A::AFF1, with a spike-in, because complete eviction destroys the readout
To see whether the fusion protein moves to IRX1 when a menin-blocking drug pushes it off HOXA, you need a dose that only partly works. At a full dose the protein leaves the DNA entirely and there is nothing left to follow.
The same thing, for a scientist
**Why the obvious experiment fails.** The only revumenib-exposed aliquot in GSE252378 (152985) has the oncoprotein off chromatin at every positive-control locus, so the 'where did it go' question has no readout there. And a **within-sample share flattens towards 1.0 under global eviction**, so a previously low locus such as IRX1 will appear to RISE as a share while falling in absolute terms - an artefact indistinguishable from an escape route [E1]. **Design.** - **Material:** KMT2A::AFF1 blasts or PDX, stratified in advance by chromatin HOXA/IRX1 status using this beat's criterion, with at least 3 HOXA-high/IRX1-low and 3 HOXA-low/IRX1-high. SEM is not a usable model: it has no IRX1 signal. - **Arms:** vehicle; a menin inhibitor at a dose titrated to a **partial** reduction in KMT2A-N occupancy at MEIS1 (target 40-70% of vehicle, verified before the primary readout is looked at); and the full dose as the eviction control. - **Calibration:** a **spike-in** in every library. GSE252378 has none, which is the single reason no absolute statement in this beat was possible. Without it the primary readout below cannot be computed at all. - **Tracks:** KMT2A-N1, KMT2A-N2, KMT2A-C1, KMT2A-C2, MEN1, H3K4me3, ENL, IgG deep enough that its promoter zero-rate is under 60%. **Primary readout, in spike-in-calibrated absolute units, not shares:** IRX1 gene-body occupancy in the partial arm versus vehicle. **Pre-declare that a rise in the IRX1 *share* with no rise in the spike-in-calibrated absolute signal is scored as renormalisation and NOT as redistribution.** That is the failure this experiment exists to avoid. **Controls that must be pre-declared:** MEIS1 and CDK6 as the within-sample positives that also define the partial dose; the eight-locus width-matched cannot-be-occupied panel as the floor; **EGR1 excluded**, because it fails its own IgG control in this assay.
| design | vehicle / partial-dose menin inhibitor / full-dose, with spike-in calibration |
|---|---|
| material | KMT2A::AFF1 primary blasts or PDX, pre-stratified by chromatin HOXA/IRX1 status |
| min n per arm | 3 |
| partial dose definition | 40-70% of vehicle KMT2A-N occupancy at MEIS1, verified before the primary readout is examined |
| primary readout | spike-in-calibrated absolute KMT2A-N at the IRX1 gene body, partial arm versus vehicle |
| prespecified artefact rule | a rise in IRX1 share with no rise in calibrated absolute signal is renormalisation, not redistribution |
| excluded control | EGR1 (fails its own IgG control in this assay) |
| unusable model | SEM (no IRX1 signal) |
tests hypotheses
[ "hyp-q147-the-irx1-address-requires-the-aff1-partner-not-the-menin-interface" ]
analyses
[ "ana-q147-irx1-fusion-occupancy-is-confined-to-kmt2a-aff1", "ana-q147-under-revumenib-hoxa-occupancy-collapses-and-the-irx1-question-becomes-unanswerable" ]
- If the oncoprotein does redistribute to IRX1 under partial menin inhibition in KMT2A::AFF1, then the drug creates the resistant state it is supposed to prevent, and an intermittent or under-dosed schedule would be worse than none. That is a dosing-schedule consequence, not a target-validity consequence, and it is currently untested.
- If it does not redistribute, the absence is only as good as the spike-in. An uncalibrated negative here would be indistinguishable from the flattening artefact, which is why the spike-in is a design requirement and not an improvement.
- Does a partial-occupancy dose exist, or is menin-inhibitor eviction effectively all-or-none at the locus level? If it is all-or-none, the redistribution question may be unanswerable in principle rather than merely unmeasured.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
In GSE252378's revumenib-exposed aliquot 152985, KMT2A-N reads 1.6x/0.8x at MEIS1 and 1.2x/1.7x at CDK6 - below the occupancy floor - while its own cannot-be-occupied floor rises to 1.030x, the highest of any KMT2A-N library in the deposit, against a deposit median of 0.206x. Under global eviction the within-sample share distribution flattens towards 1.0, so IRX1's 0.684x there is uninterpretable as either presence or absence.
q147 re-analysis of GSE252378: HOXA and IRX1 oncoprotein occupancy across all six KMT2A fusion partners, 27 samples, 133 libraries (analyses/q147-hoxa-irx1-across-partners) · Iqbal Lab oncofusion knowledge graph, beat q147 · internal analysis over deposited GSE252378 bigWigs 2026What this evidence cannot show
n = 1 drug-exposed aliquot, confounded with relapse, lineage switch and clonal replacement, with no recorded dose or exposure duration. The flattening argument is arithmetic about the normalisation rather than a measurement of the drug. Every number is a WITHIN-SAMPLE share: an interval's mean coverage divided by that same library's mean coverage over all 19,404 hg38 RefSeq promoters, so 1.0 = an average promoter in that library. GSE252378 has no spike-in, so no between-library absolute statement is licensed; the only cross-sample quantity used is a ratio of two loci measured inside one library, where the unknown per-library scale cancels.
Evidence that argues against this (1)
- The deposit paper reports that ENL persists at oncoprotein-target loci in this sample after the oncoprotein and menin have fallen to undetectable levels, so a cofactor track cannot be substituted for the unreadable KMT2A-N libraries as a workaround. Janssens DH et al. 2024