Enumeration for the subnuclear beat: every compartment queried, every count, and the papers that looked on-topic and were not used
also called subnuclear beat enumeration · compartment counted absences 2026-09-20
A record of exactly what was searched for this beat, what each search returned, and which papers were read and then set aside - so that somebody repeating the work next year repeats the decisions, not just the searches.
The same thing, for a scientist
This graph is good at bounded absence and blind to bounded rejection. This entry supplies both for the subnuclear-compartment beat, run 2026-09-20. **Counted absences**, each with the query translation read for injected `[UID]` terms and each hit title-checked: KMT2A at PML nuclear bodies **1 record** (an autophagy review); KMT2A fusion chromatin fractionation **3 records** (none fractionates a fusion); MLL/KMT2A with GPHN or gephyrin **3 records** (none reports localisation); AFF1 phase separation **0 on-topic across four alias spellings**; `(MLL OR KMT2A) AND "nuclear localization signal"` **2 records**, neither of which maps an NLS in KMT2A itself [E1]. **Bounded rejections** are listed below with reasons. The single most consequential finding of the enumeration is negative in an unexpected direction: searching `(MLL OR KMT2A) AND "nuclear localization signal"` returns two papers about *partners* (AF10/GAS41 and AF3p21), so **KMT2A's own nuclear localisation signal has never been mapped as an NLS** - the only mapping is a 1997 deletion assignment to 'a region containing the AT-hook motifs' [E2], and this graph and its sources have been repeating an unmapped signal as if it were characterised.
Counted absences, with the exact queries
| query (run 2026-09-20) | count | on-topic after title check |
|---|---|---|
(MLL OR KMT2A) AND leukemia AND (PML nuclear body OR PML bodies OR ND10 OR promyelocytic leukemia nuclear body) | 1 | 0 |
(MLL-AF4 OR MLL-AF9 OR KMT2A-AFF1 OR MLL fusion protein) AND (chromatin fractionation OR chromatin-bound fraction OR nucleoplasmic fraction) | 3 | 0 |
(MLL OR KMT2A) AND (GPHN OR gephyrin) | 3 | 0 for localisation |
AFF1 AND (phase separation OR condensate OR droplet OR liquid-liquid) | 2 | 0 |
MLLT2 AND (phase separation OR condensate OR droplet) | 0 | 0 |
(AF4 AND leukemia) AND (phase separation OR condensate OR droplet) | 2 | 0 |
(MLL OR KMT2A) AND "nuclear localization signal" | 2 | 0 for KMT2A's own NLS |
(MLL OR KMT2A) AND fusion AND (lamin B OR lamin A OR nuclear envelope) | 2 | 1 (the 2000 targeting paper) |
Every translation was inspected. None contained a [UID] term the query did not ask for, so none is the silent-rewrite failure this project has logged. No query in this beat contained a der(n) token or a parenthesised karyotype, both of which return a clean and indistinguishable 0.
Bounded rejections - read, on-topic-looking, not used
- PMID 15214844, Subcellular localization of EEN/endophilin A2 - read in full via the PMC HTML route. Used, but only for the partner-baseline entry; rejected as evidence about any fusion, because it studies wild-type EEN and no MLL-EEN construct. Would change my mind: a localisation experiment on the MLL-EEN chimera.
- PMID 11241789, AF3p21 genomic organization and cellular localization - on-topic by title. Used for partner baseline only; the fusion itself is never localised and the der(3) produced no transcript.
- PMID 36598580, Clustered PHD domains - full text read. Used for the PHD determinant entry. Rejected as evidence about subnuclear compartments, despite its imaging panel, because the constructs carry a supplied NLS and the observation (diffuse nuclear EYFP) is an aggregation control, not a compartment measurement.
- PMID 23974107, Cell context in the control of self-renewal and proliferation regulated by MLL1 - returned by the chromatin-fractionation query. Rejected: a review, no fractionation of a fusion.
- PMID 40451434 and 38822412 - returned by the same query. Rejected: SMYD3/HOXA9 and sequential-drug ALL respectively; the fractionation terms are incidental.
- PMID 37813299, Droplet Digital PCR for Oncogenic KMT2A Fusion Detection - returned by an AF4 condensate query. Rejected as a pure token collision on 'droplet', and recorded because it is a textbook instance of the alias-precision hazard.
- PMID 35817767, AUXIN RESPONSE FACTOR condensation - returned by
AFF1 AND .... Rejected: a plant paper, matched on the ARF/AF4 token family. - PMID 32917124, The role of autophagy in targeted therapy for acute myeloid leukemia - the sole hit for KMT2A at PML bodies. Rejected: a review in which PML appears as a gene name in APL, not as a nuclear body in KMT2A disease.
- Laboratories not cited. The septin-fusion group (Eguchi, Greaves) and the Joh/Kagami group are cited here. The Erfurth/Hemenway laboratory publishes repeatedly on MLL partner interactions and only one of its papers is used, because the others concern interaction rather than localisation; that is a rejection on scope, not on quality, and a future beat on the AF4/AF9/AF10 interaction web should start there.
Found, on topic, and not readable by any of the four routes
Four papers central to this beat returned no pmcid from the Europe PMC core record, so routes 2, 3 and 4 were unreachable and only the abstract could be read. They are recorded here rather than silently downgraded, and nothing in this beat asserts content beyond their abstracts.
| PMID | what its title/abstract asserts | routes tried |
|---|---|---|
| 9129043 | anti-HRX blots resolve chimeras by size in t(4;11) and t(11;19); imaging cannot distinguish them | EPMC core (no pmcid); publisher page HTTP 403 |
| 8934541 | MLL, MLL-AF9, MLL-ENL and truncated MLL nuclear by IF and cell fractionation | EPMC core (no pmcid) |
| 14603337 | AF4 and AF9 interact at 'AF4 bodies'; MLL-AF4 alters AF9's subnuclear localisation | EPMC core (no pmcid) |
| 9808577 | the 1998 anti-AF4 endogenous imaging study | EPMC core (no pmcid) |
One success worth recording for the method: PMID 15214844 was recovered by route 4. fullTextXML failed, efetch db=pmc returned a document with no <body>, and the PMC HTML page returned 178,620 bytes stripping to 49,296 characters of text. Measuring the stripped length rather than the HTML length is what distinguished this from the success-shaped stub the brief warns about.
| script | not applicable - live PubMed esearch with query translations inspected |
|---|---|
| status | run |
| result summary | eight counted absences recorded with query translations checked; nine bounded rejections listed with reasons; four papers on-topic and unreadable beyond abstract; one route-4 recovery |
| date run | 2026-09-20 |
counted absences — the numbers behind the chart
[
{
"topic": "KMT2A at PML nuclear bodies",
"raw": 1,
"on_topic": 0
},
{
"topic": "KMT2A fusion chromatin fractionation",
"raw": 3,
"on_topic": 0
},
{
"topic": "MLL/KMT2A with GPHN localisation",
"raw": 3,
"on_topic": 0
},
{
"topic": "AFF1 phase separation (4 alias spellings)",
"raw": 4,
"on_topic": 0
},
{
"topic": "KMT2A's own nuclear localisation signal",
"raw": 2,
"on_topic": 0
},
{
"topic": "KMT2A fusion and lamin/nuclear envelope",
"raw": 2,
"on_topic": 1
}
]unreadable beyond abstract
[ "9129043", "8934541", "14603337", "9808577" ]
route4 recovery
{
"pmid": "15214844",
"pmcid": "PMC1134040",
"html_bytes": 178620,
"stripped_chars": 49296,
"title": "Subcellular localization of EEN/endophilin A2, a fusion partner gene in leukaemia.",
"authors": "Cheung N, So CW, Yam JW, et al.",
"venue": "Biochem J",
"year": 2004,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}- A counted absence recorded in this format reads as authoritative and this project has twice produced a confident wrong number that way; every row here should be re-run before it is used to justify not doing an experiment.
- Using these counts to conclude a compartment is irrelevant would be an absence-of-testing error of exactly the kind this graph already documents.
- Do the untried aliases - gephyrin GPH and KIAA1385, AFF1 PBM1 - change any of these counts?
- Can the four unreadable papers be reached through an institutional route, and does the 1997 HRX paper name the cell lines it imaged?
- Has KMT2A's nuclear localisation signal genuinely never been mapped, or is it mapped in a full text that no abstract-level query reaches?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
The eight counted-absence queries listed in this entry were run against PubMed on 2026-09-20, each query translation was inspected for injected [UID] terms and none was found, and every returned record was title-checked individually with the on-topic counts recorded.
PubMed E-utilities esearch, db=pubmed: eight subnuclear-compartment queries with query translations inspected, run 2026-09-20 · Not applicable - database query · NCBI E-utilities 2026What this evidence cannot show
Every count is bounded by its symbol set. Retired and alias symbols were used where known (GPHN/gephyrin, AFF1/MLLT2/FEL/AF4) but not exhaustively - gephyrin's GPH and KIAA1385, and AFF1's PBM1, were not run. Complex-name phrases defeat subunit-symbol enumeration and were used only for the super elongation complex.
Evidence that argues against this (1)
- This project has twice produced a confident wrong absence from a well-formed query, and the bounded-absence format makes such an error read as trustworthy; the honest status of every row is 'not found by these queries', not 'does not exist'. Guo C et al. 2020
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The only published mapping of nuclear-localisation information in KMT2A is a 1997 deletion assignment to a region of the MLL N-terminus containing the AT-hook motifs; a query for KMT2A and 'nuclear localization signal' returns two records, neither of which maps a nuclear localisation signal in KMT2A itself - one concerns the partner AF10 and GAS41, the other maps a bipartite NLS in the partner AF3p21.
Chimeric MLL products with a Ras binding cytoplasmic protein AF6 involved in t(6;11) (q27;q23) leukemia localize in the nucleus · Joh T, et al. · Oncogene 1997What this evidence cannot show
Abstract only for the mapping paper; no PMCID. The two-record count is an abstract-level search and an NLS could be described in a full text without appearing in an abstract, so this is a weaker absence than the compartment counts.
Evidence that argues against this (1)
- One of the two records maps a bipartite NLS in the partner AF3p21, showing the query does retrieve NLS-mapping papers when they exist, which strengthens rather than weakens the inference for KMT2A. Hayakawa A et al. 2001
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
PMID 15214844 was unreachable by Europe PMC fullTextXML and by efetch db=pmc, which returned a document with no <body>, and was recovered from the PMC HTML page at 178,620 bytes stripping to 49,296 characters of text.
Subcellular localization of EEN/endophilin A2, a fusion partner gene in leukaemia · Cheung N, et al. · The Biochemical journal 2004What this evidence cannot show
This citation identifies the paper the retrieval was performed on; it is not cited for its content, so a claim-support scorer will score it near zero by construction. A single successful instance of route 4. The stripped-length measurement is what distinguishes a genuine recovery from a success-shaped stub, and both numbers are recorded so the check can be repeated.
Evidence that argues against this (1)
- Route 4 can return a success-shaped stub - 21,303 characters of HTML stripping to 165 characters of text - so an HTTP 200 and a large byte count prove nothing on their own. Caslini C et al. 2000
supplies the counted absence for chromatin fractionation with the query translation checked
one record intersects KMT2A leukaemia with PML nuclear bodies and it is an autophagy review
KMT2A's own NLS has never been mapped as an NLS; the 1997 region assignment is all there is
the AFF1 condensate zero is one row of this beat's enumeration
records the route-4 recovery that made the EEN partner-baseline finding readable