Bounded failure with a near-miss worth recording: a GEO series matching the topic exactly belongs to a different paper
also called PMID 34551297 bounded failure · GSE200059 attribution trap · DOT1L O-GlcNAcylation system
A paper on DOT1L being sugar-modified never names the MLL fusions it used, and it deposited nothing readable. A search of the data repository turned up exactly one dataset on precisely that topic - and it belongs to a different research group and a different paper. Attributing it would have produced a confident wrong answer.
The same thing, for a scientist
`mechanism-dot1l-oglcnacylation-stability` hedges PMID 34551297 with *"no in vivo or primary-sample data are described in the abstract, and the MLL fusions used are not named"* [E1]. The source did not resolve: no PMCID in PubMed esummary or the Europe PMC core record, `isOpenAccess: N`, `inEPMC: N`, zero GEO by elink, zero SRA/BioSample/BioProject [E2]. **The near-miss is the reason this entry exists.** A GEO keyword search on `DOT1L AND "O-GlcNAc"` returns **exactly one series, GSE200059**, titled *"Dot1L recruits OGT target chromatin to regulate histones O-GlcNAcylation"* - a title so close to the paper's subject that an agent working from topic match alone would attribute it without hesitation. It is a different study. Its `!Series_pubmed_id` is **35690146**, its seventeen contributors (Bo Xu, Can Zhang, Hui Sun and others) do not match, its submission date is April 2022 against the paper's 2021 publication, and its ten samples are **HEK293T CUT&Tag for OGT and Dot1L**, not leukaemia at all [E3]. Attributing GSE200059 to PMID 34551297 would have produced a resolution reading *the system is HEK293T* - clean, specific, and wrong about a paper whose subject is MLL-fusion leukaemia cell proliferation. **A single, exact, topic-matched hit is the most dangerous result a keyword search can return.**
The check that catches it, in one field
Every GEO series carries !Series_pubmed_id in its targ=self record when the submitters have linked a publication. Reading it takes one extra request and settles attribution outright:
GET acc.cgi?acc=GSE200059&targ=self&form=text&view=brief
!Series_pubmed_id = 35690146 <- not 34551297
!Series_contributor = Bo,,Xu <- 17 contributors, none matching
!Series_submission_date = Apr 03 2022
!Series_overall_design = Cut&Tag dataset of OGT and Dot1L in HEK293T cells
This is the mirror image of the trap in ana-unstated-system-paf-complex-species-resolved, where the linked deposition turned out to be a different experiment from the same paper. Here the deposition is a different paper on the same topic. Both fail silently and both are caught by reading the series-level record rather than the sample-level one.
The general rule: a keyword-discovered series must be attributed by !Series_pubmed_id or contributor match, never by topic. A series found through elink is attributed by construction; a series found by search is not.
What remains unknown
The entry keeps its caveat unchanged: the MLL fusions behind the O-GlcNAcylation-stability result are unnamed, and no in vivo or primary-sample data are described. This beat adds only that the unknown is not resolvable from any public repository, and that the one dataset a searcher would find is a false lead.
One substantive note that does not require the paper: the mechanism as described - a post-translational modification stabilising DOT1L protein - is fusion-agnostic on its face. If it holds, it would hold for any DOT1L-dependent leukaemia, which makes the unnamed fusion roster less consequential here than in most of this beat's sources. That is a reason to rank this bounded failure low, not a reason to close it.
| script | none - Europe PMC, NCBI elink, GEO keyword search and series-level record |
|---|---|
| status | run |
| result summary | Unresolved. No PMCID, no deposition. A single topic-exact GEO hit (GSE200059) was excluded by its Series_pubmed_id of 35690146, a non-overlapping seventeen-contributor list, an April 2022 submission date and HEK293T cell type. Attributing it would have produced a confident wrong resolution. |
| what would resolve it | Subscription access to Cell Reports 2021, or a later PMC deposit. |
| resolution class | bounded failure |
inputs
[ "PMID 34551297", "GSE200059" ]
routes tried
[ "PubMed esummary (no pmc)", "Europe PMC core (pmcid null, isOpenAccess N, inEPMC N)", "elink pubmed->gds, sra, biosample, bioproject (0 each)", "esearch db=gds 'DOT1L AND O-GlcNAc' (1 hit, GSE200059, excluded by Series_pubmed_id)" ]
near miss
{
"accession": "GSE200059",
"series_pubmed_id": "35690146",
"why_excluded": "different publication, different contributors, HEK293T not leukaemia, submitted after the paper published"
}entries affected
[ "mechanism-dot1l-oglcnacylation-stability" ]
- An OGT- or O-GlcNAc-directed strategy would perturb a nutrient-sensing modification used by thousands of nuclear proteins, so a DOT1L-stability rationale is a very narrow justification for a very broad intervention.
- If DOT1L stability is nutrient-responsive, then DOT1L inhibitor pharmacodynamics could vary with a patient's metabolic state, which is a confound no cell-line dose series would reveal.
- The near-miss dataset in this entry would have supplied a clean, specific and wrong system; any future automated provenance sweep over this graph will hit it again unless the attribution rule is enforced.
- Which MLL fusions carry the DOT1L O-GlcNAcylation stability result? Unresolvable from any public repository.
- Is DOT1L protein stability regulated the same way in a KMT2A::AFF1 cell, where DOT1L is not recruited through an AHD? If the modification is fusion-agnostic this matters less, and nobody has tested it either way.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The paper reports that DOT1L O-GlcNAcylation promotes its protein stability and MLL-fusion leukaemia cell proliferation, and its abstract does not name the MLL fusions used or describe in vivo or primary-sample data.
DOT1L O-GlcNAcylation promotes its protein stability and MLL-fusion leukemia cell proliferation. · Song T, Zou Q, Yan Y, Lv S, Li N, Zhao X, Ma X, Liu H, Tang B, Sun L · Cell Rep 2021What this evidence cannot show
Identifiers taken from the NCBI esummary record fetched on 2026-09-19. No claim is made about the paper's content beyond its abstract.
Evidence that argues against this (1)
- A mechanism acting on DOT1L protein stability rather than on fusion binding would be expected to operate across DOT1L-dependent leukaemias regardless of partner, which lowers the cost of the unnamed roster here relative to the other unresolved sources in this beat. Godfrey L et al. 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
No full text or deposition is reachable for this paper: PubMed esummary returns no PMC identifier, the Europe PMC core record returns pmcid null with isOpenAccess N and inEPMC N, and elink from the PMID to GEO, SRA, BioSample and BioProject returns zero records.
Europe PMC REST core record and NCBI elink for PMID 34551297 · Europe PMC and NCBI Entrez Programming Utilities 2026What this evidence cannot show
Run on 2026-09-19 from a single network location without subscription access. With no PMCID present, the cycle-50 rule's second clause cannot be attempted, so this is a complete rather than an ambiguous unreadable. Cell Reports is a hybrid journal and this article may be readable at the publisher.
Evidence that argues against this (1)
- Six of this beat's sixteen resolutions came from accessions printed in a data-availability statement, so an unreadable paper is denied the most productive discovery route and its deposition negative is weaker than an elink zero would be for a readable paper. Chen TQ et al. 2024
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
GEO series GSE200059, the single hit returned by a keyword search on DOT1L and O-GlcNAc and titled 'Dot1L recruits OGT target chromatin to regulate histones O-GlcNAcylation', belongs to a different publication: its Series_pubmed_id is 35690146, its seventeen contributors do not overlap with the paper's authorship, it was submitted in April 2022, and its ten samples are HEK293T CUT&Tag for OGT and Dot1L rather than leukaemia cells.
Dot1L recruits OGT target chromatin to regulate histones O-GlcNAcylation (GSE200059) · GEO submitters · NCBI Gene Expression Omnibus 2022What this evidence cannot show
Read on 2026-09-19 from the GEO accession-display text endpoint at both targ=gsm and targ=self. The exclusion rests on the series-level Series_pubmed_id field being present and pointing elsewhere; had that field been absent, contributor and date evidence alone would have made this probable rather than certain. Submitter-entered fields throughout.
Evidence that argues against this (1)
- A series can legitimately be reused across publications from the same group, so a non-matching Series_pubmed_id is not always disqualifying - here the contributor list, the submission date and the HEK293T cell type all point the same way, which is what makes the exclusion safe. Rogic S et al. 2026
- related to →How much sugar the cell has decides how much DOT1L there is - a metabolic dial on the leukaemia's mark
bounded failure; the one topic-exact GEO hit belongs to a different paper
establishes that a keyword-found series must be attributed by Series_pubmed_id, never by topic
bounded failure in the unstated-system worklist