start / graph / absence-q120-gse127507-has-no-h3k79me2-sample-for-ml-2-the-third-fusion-partner

AnalysisstrongQ120-h3k79me2-second-line-and-the-mll1-epitope

The measurement that would separate a fusion-partner effect from a lineage effect does not exist: GSE127507 deposits H3K79me2 for RS4;11 and MOLM13 and for ML-2 it deposits DOT1L, MENIN, MLL1n and an input and no H3K79me2 at all

also called no H3K79me2 in ML-2 · the third partner has no mark

Two of the three cell lines in this dataset were tested for the histone tag and the third was not. The third is the one that would tell us whether the difference we found is about the fusion partner or about the type of leukaemia.

The same thing, for a scientist

**Transcribed from the 26 sample records** [E1]: | line | fusion | DOT1L | MENIN | MLL1n | **H3K79me2** | input | |---|---|---|---|---|---|---| | RS4;11 | KMT2A::AFF1, B-ALL | d3 pair | d3 pair | d3 pair | **d3 pair** | DMSO only | | MOLM13 | KMT2A::MLLT3, AML | d3 pair | **d4** pair | d3 pair | **d3 pair** | both | | ML-2 | KMT2A::AFDN, AML | d3 pair | d3 pair | d3 pair | **ABSENT** | DMSO only | **Why the gap is exactly the wrong one.** This beat measured H3K79me2 at the fusion's own sites falling to 0.6578 in RS4;11 and not moving at all (0.8667) in MOLM13. Those two lines differ in **both** the fusion partner (AFF1 against MLLT3) and the lineage (B-ALL against AML), so the two explanations are perfectly confounded in this series. **ML-2 breaks the confound in one direction**: KMT2A::AFDN in an AML line. If H3K79me2 persisted there as in MOLM13, lineage would be the better explanation; if it fell as in RS4;11, the partner would be. **The one line that would decide it is the one line without the sample.** **This is an absence of a deposited experiment, not of an analysis.** The other seven ML-2 files exist and this beat did not download them, which is a separate and cheaper gap. No amount of re-analysis produces an H3K79me2 track for ML-2.

Facts and figures

status	run
absence claim	GSE127507 contains no H3K79me2 ChIP sample for ML-2 (KMT2A::AFDN).
date measured	2026-09-23
n records searched	26
what closes it	A menin-inhibitor H3K79me2 ChIP in ML-2 or any second KMT2A::AFDN line - wet work - or a GEO-wide query for KMT2A::AFDN H3K79me2 depositions, which was NOT run here and is the cheap first step.
why it matters	Fusion partner and lineage are perfectly confounded between RS4;11 (AFF1, B-ALL) and MOLM13 (MLLT3, AML), the only two lines with the mark, and H3K79me2 is the one channel of four that differs between them.
absence class	unclassified-not-checked
absence class note	Honest default: within GSE127507 the sample does not exist, which is certain, but NO GEO-wide or literature query has been run for an H3K79me2 ChIP in any KMT2A::AFDN line under a menin inhibitor. Until that archive query is run the class cannot be assigned - it may be never-measured or it may be measured-not-searched in another series.
absence class assigned	2026-09-23 (Q120)

How this could bite us back

A three-line panel that omits one antibody from one line looks complete in a figure legend and cannot answer the question the panel was built for. The omission is invisible unless somebody transcribes the sample list.

Still open

Has anyone deposited an H3K79me2 ChIP in a KMT2A::AFDN line under any menin inhibitor? One GEO query would answer it and this beat did not run it.
Would the seven existing ML-2 files settle anything without the mark? DOT1L, MENIN and MLL1n genome-wide in ML-2 would give a three-way ordering in a third partner, which tests the part of the ordering that DID replicate rather than the part that did not.

Every source behind this entry (1)

E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Of GSE127507's 26 samples, RS4;11 and MOLM13 each carry an H3K79me2 DMSO/VTP pair at 3 days (GSM3635697/98 and GSM3635687/88) and ML-2 carries DOT1L, MENIN and MLL1n pairs plus a DMSO input and no H3K79me2 sample; ML-2's fusion is named as MLL-AF6 (KMT2A::AFDN) in the series overall design.
GSE127507 per-sample SOFT records, 26 samples, 134,259 characters, fetched live 2026-09-23 · NCBI GEO 2019
What this evidence cannot show
An absence within one series. Another series could contain an H3K79me2 ChIP in ML-2 under a menin inhibitor and no archive query for that was run in this beat, so this bounds GSE127507 and not the literature or GEO.
Evidence that argues against this (1)
- Cancer Cell 36:660-673.e11 2019

Connected to

related to →The four-way ordering given a second cell line: menin, DOT1L and the MLL1-N channel reproduce between KMT2A::AFF1 and KMT2A::MLLT3 to within 0.04 of a share ratio, and H3K79me2 - the one channel the ordering was built around - diverges by 0.21 to 0.30 and changes places with MLL1-N
names the missing sample that would separate the fusion-partner from the lineage explanation of the one channel that does not replicate
related to →ML-2 (KMT2A::AFDN, a non-elongation partner) is the line where menin eviction is weakest (retains 37.2% of its own top windows against 20.5% and 13.0%) and where the DOT1L track moves most of any track in the beat (retains 31.6%, 13.69 percentile points below a rank-matched null at p = 3e-10) - the opposite pattern to what an AF9/ENL-free fusion would predict
ML-2's DOT1L anomaly is the flag that makes its missing H3K79me2 sample expensive
← depends onHypothesis: H3K79me2 at a KMT2A fusion's own sites survives menin inhibition when the C-terminal partner contacts DOT1L directly (MLLT3/AF9) and decays when it does not (AFF1) - so the mark's response is a property of the partner and not of menin, which predicts the opposite of the direct-contact intuition and is testable in one line
the discriminating measurement, and the reason the hypothesis cannot be scored now
← depends onThe one experiment that separates fusion partner from lineage: an H3K79me2 ChIP under a menin inhibitor in ML-2 or any KMT2A::AFDN line, alongside the MLL1n and MENIN samples that already exist for that line and can be re-binned for free
the experiment exists because the sample does not