Oncofusion

start / graph / ana-q118-molm13-is-the-only-line-with-both-inputs-and-its-mll1-n-occupancy-does-not-move

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q118-menin-at-the-human-index-fusion

MOLM13 (KMT2A::MLLT3) is the only line in GSE127507 with a matched input in BOTH conditions, and it gives the one differential-input-controlled measurement in the series - MLL1-N occupancy at its own top 0.1% of bins is unchanged (share ratio 0.9438, input 0.9542, input-corrected 0.9891) with ZERO gained bins genome-wide, while menin retains only 13.0% of its own top windows

also called MOLM13, the input-controlled line · the human MLL-AF9 arm · 0.9891

Of the three cell lines in this dataset only one has the background control for both treated and untreated samples, and it is the line carrying the same fusion as last cycle's mouse work. There the MLL1 signal does not shift at all after correcting for background, while menin keeps only one site in eight.

The same thing, for a scientist

**MOLM13 carries KMT2A::MLLT3 - the human orthologue of the murine MLL-AF9 system Q117 used - and it is the only line in GSE127507 with an input in both conditions (GSM3635689 DMSO, GSM3635690 VTP-50469).** RS4;11 and ML-2 have a DMSO input only, so **every input-corrected cross-condition statement in this beat comes from MOLM13 and nowhere else** [E1]. **The input-corrected measurement, at MOLM13 MLL1-N's own top 0.1% of 2 kb bins (universe 1,389,438, held set 1,389 bins)** [E1]: | quantity | value | |---|---:| | MLL1-N held-set share, DMSO -> VTP | 1.2549% -> 1.1843% | | **MLL1-N held-set share ratio** | **0.9438** | | input held-set share, DMSO -> VTP | 0.1443% -> 0.1377% | | **input held-set share ratio** | **0.9542** | | **input-corrected MLL1-N share ratio** | **0.9891** | | enrichment over the track's own genome median, DMSO -> VTP | 13.00x -> **14.10x** | | fraction still in its own top 0.1% | **0.6645** | | per-bin share ratio median | **0.9991**, only 6.6% below 0.5 | | **gained bins (pre-registered G1)** | **0** | | held set on an hg19 CpG island | **1,262 / 1,389 = 90.86%** (universe 2.52%) | **So in the human MLL-AF9 line, MLL1-N-reactive occupancy at its own strongest sites is unchanged to within 1.1% after input correction, and there is no gained set at all.** G1 = 0 against a threshold of 50 - the cleanest negative in the beat, and the only one with a differential-input control behind it. **Menin in the same line does leave.** At TSS windows MOLM13 menin's share ratio is 0.2513 and it retains **13.0%** of its own top 1% [E2] - which is the figure that matches Q117's murine FLAG-MLLAF9 retention of 12.8% almost exactly, in the line carrying the same fusion. **That agreement is between a menin antibody and a FLAG-tagged fusion, so it is a coincidence of magnitude and not a like-for-like comparison, and it should be quoted as such.** **The MOLM13 MENIN samples are a 4-day exposure** (`timepoint: 4 days`, GSM3635683/84) while every other sample in the series is 3 days, so MOLM13's menin figure carries one extra day of drug that its MLL1n figure does not.

Go deeper
Why the input control matters and how little it changed

The worry the control addresses is that a treated library's background distribution differs from an untreated one - through chromatin accessibility change, differential fragmentation, or copy-number drift under drug - so that a within-sample share change is partly a background change.

Measured: the input's own share at the held set moves by 4.6% (0.9542), and correcting for it moves the MLL1-N answer from 0.9438 to 0.9891. A 4.6-point correction on a 5.6-point effect - so the input correction accounts for most of the small apparent MLL1-N loss in this line, and the corrected answer is 'no change'.

That is a useful bound to carry into the RS4;11 claims, where the control does not exist. If RS4;11's input behaves like MOLM13's, the RS4;11 MLL1-N share ratio of 0.9069 corrects to roughly 0.95 and the MENIN ratio of 0.1999 corrects to roughly 0.21 - i.e. the correction is immaterial to a 5-fold effect and material to a 10% one. It is stated as a transfer and not as a measurement: the two lines have different karyotypes and the RS4;11 VTP input does not exist.

The two MOLM13 input libraries also differ substantially in total coverage - 9.46e9 against 1.22e10 coverage-bases, a 1.29-fold difference 1. That is depth, which the share statistic cancels, and it is the reason the raw totals in this series cannot be read as amounts.

Facts and figures
scriptanalyses/q118-menin-at-the-human-index-fusion/binsB.py (molm13_input_control)
statusrun
result summaryMOLM13 MLL1-N input-corrected held-set share ratio 0.9891, zero gained bins genome-wide, 90.86% of held bins on a CpG island; menin retains 13.0% of its own top 1% of TSS windows at a 4-day exposure.
cell lineMOLM13
fusionKMT2A::MLLT3 (MLL-AF9)
lineagehuman AML
drugVTP-50469 330 nM; 3 days for MLL1n/DOT1L/H3K79me2/input, 4 days for MENIN
genome buildhg19
n bins universe1389438
n held1389
antibodyanti-MLL1n Bethyl A300-086A; anti-Menin Bethyl A300-105A Lot 8
epigenetic readouthg19 CpG-island overlap of the MLL1-N held set: 1262/1389 = 90.86% vs a 2.52% universe background
shown for fusionKMT2A::MLLT3 in human MOLM13; the fusion is not distinguished from wild-type MLL1 by the MLL1n antibody
inputs
[
  "GSM3635685/86 MOLM13 MLL1n",
  "GSM3635689 MOLM13 input DMSO",
  "GSM3635690 MOLM13 input VTP",
  "GSE127507 ROSE TSS table"
]
How this could bite us back
  • MOLM13 is the line most menin-inhibitor preclinical work is done in, and in it the MLL1-N occupancy readout is flat to within 1% after input correction. Any programme using MLL1 ChIP in MOLM13 as a target-engagement assay is using an assay that does not move.
  • Zero gained bins in the input-controlled line, and 53 in the line without an input control, is the pattern a reader should be most suspicious of - and the cross-line test was run precisely because of that asymmetry. It came out positive, which means the RS4;11 gain is not simply the missing input.
Still open
  1. MOLM13's MENIN pair is the only 4-day arm in the series. Is menin's retention of 13.0% a 4-day number that would be higher at 3 days? A 3-day MOLM13 menin ChIP does not exist in this deposit.
  2. MOLM13 gives ZERO gained bins and RS4;11 gives 53, yet the 53 RS4;11-defined bins DO rise 1.77-fold in MOLM13. So MOLM13 has the effect without having it enter its own top 0.1%. Is the difference a sensitivity difference between the two libraries or a real line difference? A power calculation on the two tracks' tail distributions would say and was not run.
  3. MOLM13's MENIN samples (GSM3635683/84) were not downloaded, so the one line with both inputs has no input-corrected MENIN measurement - which is the measurement that would most strengthen the headline replication. Two files and ten minutes.
Every source behind this entry (2)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    In MOLM13 at 330 nM VTP-50469 for 3 days, MLL1-N's own top 0.1% of 2 kb bins have a within-sample share ratio of 0.9438, the matched input's share ratio at the same bins is 0.9542, the input-corrected value is 0.9891, enrichment over the track's own genome median rises 13.00x to 14.10x, 66.45% of the bins stay in the track's own top 0.1%, the per-bin share-ratio median is 0.9991 with 6.6% below 0.5, the pre-registered gained set is 0 bins, and 1,262 of 1,389 held bins (90.86%) overlap an hg19 CpG island against a 2.52% universe background.

    What this evidence cannot show

    NEITHER ANTIBODY IN THIS SERIES IS FUSION-SPECIFIC AND THE ANTIBODY NAME SAYS SO. MLL1n is anti-MLL1 N-terminus (Bethyl A300-086A, lots 5 or 6); the N-terminus is retained in KMT2A::AFF1, so that track is the sum of the fusion and wild-type MLL1 in unknown proportion. MENIN is Bethyl A300-105A Lot 8; menin binds the fusion, wild-type MLL1 and non-KMT2A partners. Q117 could attribute its numbers because its fusion was FLAG-tagged; nothing here can. Every occupancy statement in this beat is about MLL1-N-reactive material or menin-occupied sites, never about the fusion. No scale factor exists for any sample. The depositors state that a Drosophila spike-in was run on one batch, that fly/human ratios were 'found to be small', and that NO spike-in normalisation was applied; no number for that measurement is reported and none is deposited. Every cross-condition figure here is therefore a WITHIN-SAMPLE share or a within-sample rank, which cancel an unknown per-sample scale factor exactly, and never a fold change. n = 1 per antibody per condition per line - there is no replicate anywhere in GSE127507 - so no number here carries a spread and none is precise. MOLM13 is a myeloid KMT2A::MLLT3 line, NOT the index fusion and NOT B-lineage. The input correction is a ratio of two within-sample share ratios and assumes the input is a valid background model for the ChIP, which is the standard assumption and not a verified one. A share ratio near 1.0 is consistent with total retention AND with a uniform global loss; the unnormalised whole-file totals of this pair are 3.611e9 (DMSO) and 3.867e9 (VTP) coverage-bases, which cannot be read as amounts because library depth is not controlled.

    Evidence that argues against this (1)
    • The two MOLM13 input libraries differ 1.29-fold in total coverage (9.46e9 vs 1.22e10 coverage-bases). A within-sample share cancels that, but a differential-input CORRECTION built from two libraries of very different depth is a noisier instrument than it looks, and there is no replicate to bound it. Iqbal Lab oncofusion knowledge graph 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    At TSS windows in MOLM13, menin's within-sample share ratio at its own top 1% is 0.2513 and 13.0% of those windows remain in the track's own top 1% under VTP-50469, against the murine FLAG-MLLAF9 figure of 12.8% at its own top 0.1% of genome-wide bins.

    What this evidence cannot show

    The MOLM13 MENIN pair is a 4-DAY exposure and every other sample in the series is 3 days, so this figure carries one extra day of drug. The comparison to Q117's 12.8% is between a MENIN antibody and a FLAG-tagged fusion, over different universes (21,461 promoter windows at top 1% versus 1,389,438 genome-wide bins at top 0.1%) and different species - it is an agreement of magnitude and not a like-for-like replication. n = 1 per antibody per condition per line - there is no replicate anywhere in GSE127507 - so no number here carries a spread and none is precise.

    Evidence that argues against this (1)
    • Because the universes and the top fractions differ, the numerical closeness of 13.0% and 12.8% carries less information than it appears to. This beat's genome-wide RS4;11 MENIN figure of 10.58% at top 0.1% is the construction-matched comparison and it is the one the replication entry quotes. Iqbal Lab oncofusion knowledge graph 2026
Connected to