Two of the nineteen non-AFF1 samples are menin-inhibitor treated, so the specificity arm is effectively seventeen - and removing them does not rescue the core
also called q152 C2 treated-sample check
The claim 'switched on in KMT2A::AFF1 and in none of the other nineteen patients' counts two patients who were on a menin inhibitor, a drug that strips the oncoprotein off DNA. Finding nothing in a treated patient is almost free, so the comparison is really against seventeen patients. Redoing it that way does not change the answer.
The same thing, for a scientist
Pre-registered escape condition C2. Samples 148752 and 152985 are KMT2A-ENL blasts flagged `treated: true` in this project's q147 per-sample table; one of them (152985) has a 2.17% pass rate, the second lowest in the deposit, while 148752 passes 11.1%. Requiring absence in only the 17 untreated non-AFF1 samples raises the AFF1-only set from 332 to 342 and leaves the recurrent core at exactly 17. Re-running the label permutation over the 25-sample pool gives a null of mean 11.9 (95th percentile 25) and p_upper 0.2163 - the negative gets slightly stronger, not weaker. C2 does not fire as a rescue; it fires as a labelling correction.
Why this had to be checked and what it cost
Absence of occupancy in a drug-treated blast is not evidence that the promoter is partner-specific, because the drug removes the oncoprotein from its own sites. Any claim of the form 'and 0 of 19' should therefore be written 'and 0 of 17 untreated, plus 2 treated'. The magnitude of the effect here is small because dropping samples from the comparator can only enlarge the exclusive set: 332 -> 342, a 3.0% increase, and the >=3-of-8 core is unchanged at 17 with the same membership 1.
The correction to make in the earlier entries
The '0 of 19 non-AFF1' phrasing appears in this beat's own dispatch brief and is the natural reading of the q151 calibration. It should read 0 of 17 untreated non-AFF1 samples and 2 menin-inhibitor-treated samples in which absence is uninformative. Nothing numerical changes; the strength of the specificity claim does.
| script | analyses/q152-the-aff1-only-recurrent-promoter-core/step2.py |
|---|---|
| status | run |
| result summary | Comparator arm is 17 informative samples, not 19. AFF1-only 332 -> 342, core unchanged at 17, permutation p 0.1119 -> 0.2163. |
measurements
[
{
"quantity": "n_informative_comparator_samples",
"subject": "the non-AFF1 arm of the AFF1-only criterion",
"value": 17,
"units": "samples",
"n": 19,
"universe": "the 19 non-KMT2A::AFF1 samples of GSE252378",
"null": null,
"note": "148752 and 152985 are menin-inhibitor treated; absence of occupancy in a treated blast is nearly free"
},
{
"quantity": "count_partner_exclusive_promoters_untreated_comparator",
"subject": "promoters occupied in >=1 of 8 AFF1 and 0 of the 17 UNTREATED non-AFF1 samples",
"value": 342,
"units": "promoters",
"n": 19404,
"universe": "19,404 refGene 2 kb promoters, two-antibody 2.0x plus floor rule",
"null": null,
"note": "up from 332 with all 19; the recurrent core is unchanged at 17"
}
]- Should treated samples be excluded from every partner-specificity comparator in this project by default, and does any existing claim depend on their inclusion?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
Samples 148752 and 152985 of GSE252378 are KMT2A-ENL and menin-inhibitor treated with promoter pass rates of 11.1% and 2.17%; requiring absence in only the 17 untreated non-AFF1 samples gives 342 AFF1-only promoters (up from 332) and the identical 17-promoter recurrent core, and the label-permutation null over the resulting 25-sample pool is mean 11.9 with 95th percentile 25 and p_upper 0.2163.
Q152 escape-condition C2, treated-sample removal and re-permutation · Iqbal Lab oncofusion knowledge graph, beat Q152 · this project 2026What this evidence cannot show
Treatment status comes from this project's q147 per-sample table, which was built from the deposit metadata; it was not re-read from GEO in this beat.
Evidence that argues against this (1)
- One of the two treated samples (148752) passes 11.1% of promoters, close to the 10.75% cohort median, so its occupancy has not collapsed globally and its absences may be as informative as any untreated sample's. Treating both as uninformative on the basis of a metadata flag is a simplification, and the correct treatment may be sample-specific. NCBI Gene Expression Omnibus 2024
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In this deposit, occupancy at the fusion's own promoters collapses under a menin inhibitor, which is why absence in a treated sample carries little information.
MLL oncoprotein levels influence leukemia lineage identities · Janssens DH, Duran M, Otto DJ, Wu W, Xu Y, Kirkey D, Mullighan CG, Yi JS, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · Nat Commun 15:9341 2024What this evidence cannot show
The paper's menin-inhibitor arm is the basis for the claim; the magnitude of the collapse in these two specific libraries was not re-measured here, only their global pass rates (11.1% and 2.17%).
Evidence that argues against this (1)
- This project's own q117 and q141 work found that a menin inhibitor leaves roughly 18-25% of fusion occupancy in place, so a treated sample is not a blank; a promoter genuinely occupied in an untreated blast of the same partner could still be detectable after treatment, which would make an absence there informative after all. Iqbal Lab oncofusion knowledge graph et al. 2026
the comparator arm is 17 informative samples, and correcting it strengthens the negative
the two comparator samples are menin-inhibitor treated, which makes their absence uninformative