The AHD inhibitor's deposited arm is MOLM-13 and its in vivo arm is nowhere: a half-resolution that names the limit exactly
also called GSE151660 provenance · AF9/ENL-DOT1L inhibitor system · AHD inhibitor in vivo model
Two entries note that a small-molecule AHD inhibitor paper never says which fusion its animal model carried. The deposited sequencing is six samples of one human leukaemia line, MOLM-13, which carries MLL-AF9 - and the animal experiment left no deposited data at all, so that half of the question stays open.
The same thing, for a scientist
`import-class-ahd-dependent-mechanism` and `risk-partner-import-distorts-target-selection` both hedge PMID 34373735 with *"the in vivo model is not specified by fusion in the abstract"* [E1]. GEO series **GSE151660** resolves the cell-culture arm and, by its absence, sharpens the caveat rather than removing it. Six *Homo sapiens* samples, all `cell line: Molm-13`, source name *"Leukemia cells Molm-13"* - three untreated and three at `5 uM inhibitor`, titled `Treated_Con5_1/2/3` [E2]. MOLM-13 carries KMT2A::MLLT3 and is myeloid, so the deposited arm is a **two-axis import**: partner MLLT3 rather than AFF1, lineage myeloid rather than B-lymphoid, with species human. **The in vivo arm is not in the deposition**, and the full text describes tumour-growth work without a deposited sequencing counterpart [E3]. This is the cleanest illustration in the beat of the standing bound: *of the deposited sequencing, never of the study*. The correct rewrite of the caveat is not *the model is MOLM-13* but *the in vitro arm is MOLM-13 and the in vivo model remains unresolved because it generated no deposited data*.
What resolved and what did not
| Arm | Resolved? | Evidence |
|---|---|---|
| Transcriptional response to the inhibitor | Yes - MOLM-13, human, 5 uM, n = 3 per arm | GSE151660 2 |
| Which fusion the cultured cells carry | Yes - KMT2A::MLLT3 (MOLM-13) | GSE151660 2 |
| Which fusion the in vivo model carries | No - no deposited sequencing exists for it | 3 |
| Whether a KMT2A::AFF1 model was used anywhere | No - absent from the deposition; unknown for the paper | 23 |
The full text names MOLM-13 18 times and MV4;11 8 times, so a second human line is in play in the paper that is absent from the deposition entirely 3. That asymmetry is itself the lesson: a deposition can under-report a paper's systems as easily as an abstract can.
Why this one matters for the menin interface specifically
import-class-ahd-dependent-mechanism exists to mark a whole class of results that reach this graph through the ANC1 homology domain, a module MLLT3 and MLLT1 carry and AFF1 does not. An AHD-competitive inhibitor tested in MOLM-13 is being tested in a cell whose fusion contains the target interface. There is no reason to expect the same compound to matter in KMT2A::AFF1, where the fusion has no AHD of its own and would have to reach one in trans.
So this resolution does not weaken an entry the way most do; it confirms the entry's own thesis with a specific accession. risk-partner-import-distorts-target-selection can now name GSE151660 as a worked instance rather than a worry.
| script | none - GEO accession-display text endpoint plus elink |
|---|---|
| status | run |
| result summary | 6 of 6 deposited samples are human MOLM-13 (KMT2A::MLLT3, myeloid) at 5 uM inhibitor versus untreated. The in vivo model's fusion remains unresolved: no sequencing was deposited for it. The full text also names MV4;11, which is absent from the deposition. |
inputs
[ "PMID 34373735", "GSE151660", "PMC8344022" ]
counts — the numbers behind the chart
{
"samples_total": 6,
"samples_human": 6,
"samples_treated": 3,
"inhibitor_uM": 5,
"cell_lines_deposited": 1,
"cell_lines_named_in_full_text": 2,
"in_vivo_samples_deposited": 0
}entries affected
[ "import-class-ahd-dependent-mechanism", "risk-partner-import-distorts-target-selection" ]
import axes
[ "fusion: KMT2A::MLLT3 not KMT2A::AFF1", "lineage: myeloid not B-lymphoid" ]
- An AHD-competitive agent prioritised for infant KMT2A::AFF1 disease from MOLM-13 data is being prioritised from a cell whose fusion contains the target interface, in a disease whose fusion does not; the mechanism does not transfer by default.
- 5 uM is a high exposure for a protein-protein interaction inhibitor, so the transcriptional signature in this deposition may include substantial off-target effects that would not be reproduced at a tolerated dose.
- The AHD is used by normal AF9 and ENL in healthy haematopoiesis, so a compound that genuinely blocks it inherits an on-target normal-progenitor toxicity that a single-line transcriptional experiment cannot reveal.
- Which fusion did the in vivo model carry? The paper can answer it and the deposition cannot; this is the single clearest instance in the beat of a question that needs the Methods section rather than a repository.
- Does an AHD-competitive compound do anything at all in a KMT2A::AFF1 line, where the fusion carries no AHD? Nothing in this deposition or in the graph answers it.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The AHD inhibitor paper reports that a small-molecule inhibitor of AF9/ENL-DOT1L/AF4/AFF4 interactions suppresses malignant gene expression and tumour growth, and its abstract does not specify the fusion carried by the in vivo model.
Small-molecule inhibitor of AF9/ENL-DOT1L/AF4/AFF4 interactions suppresses malignant gene expression and tumor growth. · Wu F, Nie S, Yao Y, Huo T, Li X, Wu X, Zhao J, Lin YL, Zhang Y, Mo Q, Song Y · Theranostics 2021What this evidence cannot show
Identifiers taken from the NCBI esummary record fetched on 2026-09-19. Full text read from Europe PMC PMC8344022 (open access) on the same date. The abstract-level claim is the one the graph's caveat was anchored to.
Evidence that argues against this (1)
- DOT1L has three separate AF9-binding sites and structure-guided mutations produce graded rather than all-or-none loss of recruitment, so a single competitive small molecule is unlikely to abolish the interaction cleanly and an apparent phenotype may reflect partial occupancy plus off-target activity. Kuntimaddi A et al. 2015
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
GEO series GSE151660 contains six Homo sapiens samples, all annotated 'cell line: Molm-13' with source name 'Leukemia cells Molm-13', comprising three untreated and three treated at 5 uM inhibitor.
Expression changes after Inhibition of AHD in Molm-13 cells (GSE151660) · GEO submitters · NCBI Gene Expression Omnibus 2021What this evidence cannot show
Read on 2026-09-19 from the GEO accession-display text endpoint. Submitter-entered; the treatment field is misspelled 'agebt' rather than 'agent' in all six records, another instance of uncurated submitter metadata. The compound is identified only as 'inhibitor' with no name or catalogue number, so the deposition fixes the cell system but not the chemical matter.
Evidence that argues against this (1)
- Six samples at a single 5 uM concentration with no dose series and no second cell line is a thin transcriptional arm, and 5 uM is high enough that off-target transcriptional effects cannot be excluded from the deposition alone. Rogic S et al. 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
No deposited sequencing corresponds to the paper's in vivo tumour-growth experiment: the only GEO series linked to the paper is the six-sample MOLM-13 culture series, and elink from this PMID to SRA, BioSample and BioProject returns zero records.
NCBI elink pubmed to gds, sra, biosample and bioproject for PMID 34373735 · NCBI Entrez Programming Utilities 2026What this evidence cannot show
Queried on 2026-09-19. A bounded absence with a known blind spot: elink from PubMed to SRA, BioSample and BioProject returned zero for every one of the 26 papers this beat tested, including papers with confirmed GEO series, so that half of the query is uninformative. The GEO half is informative because elink to db=gds did return GSE151660 for this PMID. An in vivo arm could also be deposited without being linked, or not deposited at all; these cannot be distinguished here.
Evidence that argues against this (1)
- Six of this beat's sixteen resolutions were found only through accessions printed in a data-availability statement rather than through elink, so an elink negative is weak evidence of no deposition and the in vivo arm could exist under an unlinked accession. Wu F et al. 2021
- supports →Import class 2: AHD-dependent mechanism, which cannot transfer because KMT2A::AFF1 has no AHD
supplies a named accession for the import class: GSE151660 is MOLM-13, a fusion that carries the AHD being targeted
worked instance of partner import distorting an intervention rationale
- related to →KMT2A::MLLT3 (MLL-AF9), t(9;11)(p21.3;q23.3)
the deposited arm is MOLM-13, KMT2A::MLLT3
an AF9/ENL-DOT1L interaction inhibitor tested in a MLLT3 fusion
- ← analogous toThe lncRNA-DOT1L-LAMP5 axis was measured on H3K79 in MOLM-13 alone: eight human samples, one fusion, no index-fusion arm
both depositions are narrower than their papers, omitting arms the graph cites
resolved source in the unstated-system worklist