The cheapest unrun test in this beat's queue: GSE260742 deposits an H3K79me2 ChIP under the DOT1L inhibitor SGC0946 in the same murine MLLAF9 cells as its FLAG and MLL1 tracks, which removes the catalysis directly instead of indirectly and would separate 'the mark is stable' from 'residual DOT1L is sufficient' - one download
also called the SGC0946 arm · one download away
There is a second drug in the same public dataset that blocks the tag-writing enzyme directly rather than through menin. Comparing the two would say whether the tag survives because it is stable or because enough enzyme is left, and it is a single file.
The same thing, for a scientist
**The question this beat opened and cannot close.** In MOLM13, DOT1L occupancy at the fusion's own sites falls to 0.2773 under a menin inhibitor and H3K79me2 there does not move (0.9980). Two readings: the mark is intrinsically stable over three days, or 22% residual DOT1L occupancy is catalytically sufficient. A menin inhibitor cannot separate them because it removes the writer indirectly and incompletely. **A DOT1L inhibitor can.** SGC0946 blocks the methyltransferase itself, so residual occupancy is irrelevant. **GSE260742 deposits an SGC0946 arm in the same murine FLAG-MLLAF9 cells**, alongside the VTP50469 and DMSO arms this beat used, with H3K79me2 read by Abcam ab3594 [E1]. Its FLAG counterpart (GSM8123360) is already binned in this project from Q117. **The pre-registered prediction is in `hyp-q120`'s MECHANISM ARM falsifier**: H3K79me2 at the FLAG-tagged fusion's own top 0.1% of bins below 0.50 under SGC0946 while staying above 0.85 under VTP50469 supports residual-sufficiency; both drugs leaving the mark high refutes both readings and says the mark is not DOT1L-limited on this timescale at all. Bands 0.45-0.55 and 0.80-0.90. **And it bears on a number this graph records as never measured.** `absence-h3k79me2-decay-half-life-never-measured` bounds the absence with a dated, controlled query. SGC0946 at 72 h in a dividing culture is a one-point decay measurement with the writer fully inhibited - not a half-life, but the first bound on one from deposited data. **Why it was not run here.** This beat had two arms and a session; the SGC0946 arm is a third question. It is recorded as a named, costed, unrun analysis rather than as a caveat, because a caveat naming a gap that costs one download is a task.
| script | not written |
|---|---|
| status | proposed |
| cost | one to three downloads, binned and deleted, plus one script; no wet work |
| system | murine FLAG-MLLAF9, GSE260742 |
| drug | SGC0946 5 uM 72 h against VTP50469 500 nM 48 h and DMSO |
| readout | H3K79me2 aggregate within-sample share ratio at the FLAG-DMSO top 0.1% of mm10 2 kb bins, under each drug |
| what it would settle | whether H3K79me2's persistence under a menin inhibitor reflects an intrinsically stable mark or catalytically sufficient residual DOT1L |
| prerequisite | transcribe the murine H3K79me2 sample accessions from the GSE260742 SOFT records; they were not identified in this beat |
| epigenetic readout | H3K79me2 under direct DOT1L catalytic inhibition, in a tagged-fusion system |
| mark written | H3K79me2 |
| writer of that mark | gene-dot1l |
| shown for fusion | MLL-AF9 in a murine FLAG-tagged model; the data exists and is unanalysed |
| absence class | measured-not-analysed |
| absence class note | The data is deposited and public in GSE260742 and this analysis has never been run on it. Compute, not wet work - one to three downloads. |
| absence class assigned | 2026-09-23 (Q120) |
- If SGC0946 also leaves H3K79me2 high at 72 h, then H3K79me2 is not a usable pharmacodynamic marker for either drug class on a clinically relevant timescale, and pinometostat's own development used exactly that readout.
- What are the accessions of the murine MLLAF9 H3K79me2 samples in GSE260742? Not transcribed in this beat and they must be read from the records, not guessed from the numbering.
- Does the SGC0946 arm's 72 h exposure against the VTP arm's 48 h confound the comparison? It is the same asymmetry this beat measured in the FLAG and MLL1 arms, and here the two drugs are the things being compared, so unlike ARM 2 the mismatch does NOT cancel.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
GSE260742 deposits, in murine MLLAF9 cells, a three-condition set per antibody - DMSO, SGC0946 5 uM for 72 h, and VTP50469 500 nM for 48 h - including FLAG (GSM8123359/60/61) and MLL1 (GSM8123395/96/97), with H3K79me2 read by Abcam ab3594 per the paper's key-resources sheet and SGC0946 sourced as Sigma-Aldrich SML1107; the FLAG SGC0946 track (GSM8123360) is already binned at 2 kb in this project's scratch from Q117.
What this evidence cannot show
The specific accession of the murine MLLAF9 H3K79me2 SGC0946 sample was NOT identified in this beat - only the MLL1 and FLAG triplets were transcribed from the 158 records, and the H3K79me2 arm's accessions were not. Whoever runs this must transcribe them from the records first rather than inferring them from the numbering, which is the mistake this project has made before. CST #5427 anti-H3K79me2 reads a histone MARK, i.e. a consequence. It cannot distinguish halted deposition from active demethylation from replicative dilution, and it cannot attribute any of them to the fusion. Abcam ab3594 is a DIFFERENT anti-H3K79me2 polyclonal from GSE127507's CST #5427, so a mouse-to-human comparison of the mark crosses reagents.
Evidence that argues against this (1)
the MECHANISM ARM falsifier: direct DOT1L catalytic inhibition separates 'the mark is stable' from 'residual DOT1L is sufficient'
names the deposited data that would give the first bound on the mark's decay with the writer catalytically inhibited, for one download
the experiment that would choose between that mechanism entry's three readings
the one-download framing does not survive contact with the batch structure: the SGC0946 arm is in a different experiment from the timecourse, and criterion K4 shows the two batches are not on one scale (24 h timecourse 0.6790 against 48 h main arm 0.7734 at the same set)