GSE252378 deposits no transcriptional readout: all 171 samples are library_strategy OTHER and the metadata never says RNA, so every programme claim from this deposit is occupancy and none is expression
The dataset everyone is using to argue about which genes these leukaemias switch on does not measure gene activity at all. It measures where a protein sits on DNA. Those are different things and the dataset cannot tell you the first one.
The same thing, for a scientist
Counted on the deposit's own SOFT family file: **171 of 171 GSMs carry `!Sample_library_strategy = OTHER`**, and the file contains **zero** case-insensitive matches for `RNA-seq` or `transcriptome`. `!Series_type = Other`. So the accession contains AutoCUT&RUN and CUT&Tag occupancy and nothing else. This matters because the questions the deposit is used for are phrased in expression language. The **HOXA9/IRX1 subgroup split** that motivates the HOXA-versus-IRX1 line of work is an expression dichotomy derived by Kallisto/tximport/DESeq2 over two RNA-seq cohorts, with **no chromatin experiment of any kind**. The deposit's own paper speaks of oncoproteins that *"preferentially activate"* an ALL or an AML programme - activation is a transcriptional word applied to an occupancy measurement. **The operational rule for anything built on GSE252378: write 'occupied', never 'expressed'; write 'the oncoprotein is on these promoters', never 'these genes are on'.** An occupancy negative does not refute an expression claim and an occupancy positive does not confirm one.
What would close it, and at what cost
The class here is never-measured: no transcriptional instrument was pointed at these aliquots. Closing it needs either new RNA from the same banked material - which is a wet-lab request to four institutions - or a different accession with paired occupancy and RNA in KMT2A-rearranged primary samples.
The single most valuable missing measurement in this whole line of work is A73051's own IRX1 and HOXA9 transcript level. It is the one depth-adequate HOXA-negative, IRX1-negative blast, and one RNA measurement would separate 'the expression subgroup' from 'the occupancy subgroup' in a single sample. It is not in this deposit and no re-analysis can produce it.
Why this is filed as a concept and not left in a caveat
Three separate beats in this corpus have now scored programmes in this deposit, and each has had to restate the same warning in its own caveats field. A warning that lives only in caveats protects the reader who opens the entry and does nothing for the reader who copies the claim. This is the shared statement, with its count, so the next beat links instead of re-deriving.
| absence class | never-measured |
|---|---|
| deposit | GSE252378 |
| n gsms | 171 |
| n library strategy other | 171 |
| n rna or transcriptome mentions in soft | 0 |
| series type | Other |
| what would close it | new RNA from the same banked aliquots at Texas Children's, St. Jude, Fred Hutch and Seattle Children's, or a different accession with paired occupancy and RNA in primary KMT2A-rearranged samples |
| highest value single missing measurement | A73051's own IRX1 and HOXA9 transcript level |
| companion accession not checked | GSE159608 |
| status | run |
measurements
[
{
"quantity": "gsms_with_library_strategy_OTHER",
"subject": "GSE252378 deposited samples",
"value": 171,
"units": "samples",
"n": 171,
"universe": "every !Sample_library_strategy line in the GSE252378 SOFT family file",
"null": null,
"note": "no other value appears; 0 mentions of RNA-seq or transcriptome in the file"
}
]- Does the companion accession GSE159608 carry any RNA for these patients? It was not checked in this beat and it is one elink away.
- Is there any deposit anywhere with paired oncoprotein occupancy and RNA in the same primary KMT2A-rearranged aliquot? Without one, the occupancy and expression literatures on this disease cannot be joined at all.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
All 171 GSE252378 samples carry !Sample_library_strategy = OTHER and the deposit's SOFT family file contains no mention of RNA-seq or transcriptome.
GSE252378, Gene Expression Omnibus - KMT2A oncoproteins induce epigenetic resistance to targeted therapies; 171 sample records, Series_pubmed_id 39472576, read from the SOFT family file 2026-09-24 · Janssens DH, Ahmad K, Henikoff S · NCBI Gene Expression Omnibus 2024What this evidence cannot show
Counted on a locally cached copy of the SOFT family file, fetched in an earlier cycle rather than today; the count was re-run this cycle on that cached file. A cached metadata file could in principle be stale if the submitters added samples, which would change the denominator but not the direction.
Evidence that argues against this (1)
- The paper reports 36 profiled MLL-rearranged samples and cites a second accession (GSE159608) for some CUT&Tag channels, so the absence of RNA is a statement about GSE252378 and not about the study's whole data footprint. Whether GSE159608 or any companion accession carries RNA was not checked here. Janssens DH et al. 2024
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The HOXA9/IRX1 subgroup split is an expression result from two RNA-seq cohorts with no chromatin experiment.
HOXA9/IRX1 expression pattern defines two subgroups of infant MLL-AF4-driven acute lymphoblastic leukemia. · Symeonidou V, Ottersbach K · Experimental Hematology 93:38-43.e5 (issue Jan 2021; epub 2020-10-15) 2021What this evidence cannot show
A re-analysis of two previously published RNA-seq data sets, so it inherits their cohorts and their batch structure; the mutual exclusivity is a clustering result, and a clustering offering two labels assigns every sample one of them.
Evidence that argues against this (1)
- Because the split is transcriptional and this deposit is not, the two literatures cannot be made to agree or disagree. That is not a criticism of either; it is the reason a third measurement - paired RNA and occupancy in the same aliquot - is the only thing that would settle it. Janssens DH et al. 2024
the deposit-level reason occupancy and expression cannot be joined here
bounds that measurement to occupancy language