Oncofusion

start / graph / absence-q125-no-transcriptional-readout-is-deposited-for-the-ntvko-cultures-so-silenced-means-marks-and-not-expression

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q125-the-h2ak119ub-arm-of-the-pcgf1-knockout

This cycle's word 'silenced' means two repressive marks up and one active mark down at 2 kb resolution - GSE260742 deposits no RNA-seq, no nascent transcription and no reporter for either NTvKO genotype, so whether the vacated sites stop being transcribed is unmeasured in the experiment that established they are marked for silencing

also called silenced means marked, not measured silent · no RNA in the NTvKO arm

We can see the chemical 'off' labels being added to the DNA the drug clears, but nobody measured whether those stretches actually stopped being read. The word 'silenced' in this work means labelled for silencing, not observed silent.

The same thing, for a scientist

**The gap is in the same experiment as the result, which is what makes it worth an entry.** This cycle measured H2AK119ub and H3K27me3 rising and H3K4me3 falling, monotonically in fusion-eviction depth, at the bins a menin inhibitor clears. **All three are marks.** Across GSE260742's 158 sample records every one is `!Sample_library_strategy = ChIP-Seq` with `!Sample_library_selection = ChIP`; **no RNA-seq, PRO-seq, GRO-seq, TT-seq, nascent or reporter sample exists for the NTvKO cultures or for any other arm of the series** [E1]. **So the inference from mark to transcription is imported, not measured here**, and it is imported into a system where the corpus already knows the arrow is contested: Sparbier et al. report that menin inhibition *paradoxically ACTIVATES* bivalent genes by disrupting the KMT2A/B-polycomb balance, so H3K27me3 gain at a polycomb-competent locus does not license a transcriptional prediction without measurement [E2]. **And the H3K4me3 result makes the gap sharper rather than softer.** H3K4me3 falls at the most-evicted quartile (0.9491) and RISES at the least (1.2025). **A reader who takes 'silenced' literally will predict transcriptional loss at Q1 and gain at Q4 in the same cells under the same drug** - a strong, specific, entirely untested claim that this cycle's language invites and its data cannot support [E1].

Go deeper
What would close it, in cost order
  1. The paper's own RNA-seq, if it deposited any in a different series. PMID 41634341's data availability statement was not re-read in this beat for a second accession; elink pubmed->gds on that PMID returns whatever series the paper registered and has not been run for this purpose. One request.
  2. Any published RNA-seq of murine MLL-AF9 under VTP-50469 at 48 h, joined to these bins by coordinate. Not the same cells, so it bounds rather than closes.
  3. RNA-seq of the NT and C1 cultures under the same three conditions. That is wet work in the depositing laboratory and is the only thing that closes it properly.
Facts and figures
absence classnever-measured
absence class noteFor the NTvKO cultures this is never-measured in GSE260742: every one of its 158 samples is ChIP-Seq and no transcriptional readout of any kind is deposited. It may be MEASURED-NOT-SEARCHED at the level of the paper - `elink pubmed->gds` on PMID 41634341 has not been run for a second accession and the data-availability statement was not re-read - and that one request is what would reclassify it.
absence class assigned2026-09-23 (Q125)
searched!Sample_library_strategy and !Sample_library_selection on all 158 GSE260742 records
what would close it
[
  "elink pubmed->gds on PMID 41634341 for a second deposited series",
  "published RNA-seq of murine MLL-AF9 under VTP-50469 at 48 h, joined by coordinate",
  "RNA-seq of the NT and C1 cultures under DMSO / SGC0946 / VTP-50469"
]
How this could bite us back
  • Every therapeutic argument in this cycle runs through the word 'silenced'. If the vacated sites acquire polycomb marks and keep being transcribed - which is what Sparbier's paradoxical activation would look like at these loci - then the whole mechanism is a chromatin epiphenomenon and the resistance predictions built on it are wrong.
Still open
  1. Does PMID 41634341 deposit a second, transcriptional series? One elink request would say, and it would either close this absence or make it much stronger.
  2. If transcription were measured, would it follow the gradient or the aggregate? The two make opposite predictions at Q1 and Q4, which is an unusually sharp test for a question about chromatin and expression.
Every source behind this entry (2)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    All 158 sample records of GSE260742 carry !Sample_library_strategy = ChIP-Seq and !Sample_library_selection = ChIP; no RNA-seq, nascent-transcription or reporter sample is deposited for the NTvKO cultures or for any other arm, so the transcriptional consequence of the mark changes measured in this cycle is unmeasured in this series.

    What this evidence cannot show

    Bounds the DEPOSIT. The depositing laboratory may hold or have published RNA-seq of these cultures elsewhere: the paper's data-availability statement was not re-read in this beat for a second accession, and `elink pubmed->gds` on PMID 41634341 was not run for this purpose. So this is 'not in GSE260742', not 'nowhere', and the first step named above would sharpen it in one request.

    Evidence that argues against this (1)
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The direction from polycomb mark to transcription at these loci is contested in the published literature: Sparbier et al. report that targeting menin disrupts the KMT2A/B and polycomb balance to paradoxically ACTIVATE bivalent genes, so a gain of H3K27me3 under a menin inhibitor does not by itself license a prediction of transcriptional loss.

    Targeting Menin disrupts the KMT2A/B and polycomb balance to paradoxically activate bivalent genes. · Sparbier CE, Gillespie A, Gomez J, et al. · Nature Cell Biology 2023
    PMID 36635503DOI
    What this evidence cannot show

    Their system and locus set are not this beat's: bivalent promoters rather than the fusion's top-0.1% occupancy set stratified by eviction depth, and this beat did not compute the overlap. Cited to establish that the inference is contested, not that it is wrong.

    Evidence that argues against this (1)
Connected to