The KMT2A::MLLT1 mechanism sweep: what was searched, what was read and not used, and the three retrieval traps that would have produced confident wrong answers
also called MLLT1 mechanism enumeration · ENL search audit · t(11;19) bounded rejection register
This is the audit trail for the mechanism half of this beat: every search that was run with its result, every paper that looked relevant and was set aside with the reason, and the three searches that returned a clean, believable zero for the wrong reason. It exists so that someone re-running this beat next year repeats the decisions rather than only the searches.
The same thing, for a scientist
Bounded enumeration for the mechanism, epigenetics, localisation and outcome sweep of KMT2A::MLLT1, run 2026-09-20. Records (1) the executed queries with counts and query translations, (2) papers found on topic and deliberately not written up, with reasons and with what would change the decision, (3) papers that are on topic and could not be read after all four PMC routes plus the publisher landing page, and (4) three retrieval traps encountered live: a t(11;19) query rewritten to 11,19[UID] returning a clean 0; a fusion-name condensate query that misses the only primary fusion-context paper because its abstract says 'the fusion of ENL with MLL'; and a PMC HTML page that returns 21,303 characters of markup which strip to 165 characters of text, i.e. a stub that passes a length check on the raw HTML.
Queries executed, with counts
All run against PubMed E-utilities on 2026-09-20, query translations inspected 1.
| query | count | note |
|---|---|---|
MLLT1 AND DOT1L | 5 | small and clean |
(MLL-ENL OR MLL::ENL OR KMT2A-MLLT1) AND (H3K79 OR DOT1L) | 9 | |
MLLT1 AND (super elongation complex OR P-TEFb OR CDK9) | 9 | |
ENL[ti] AND YEATS[tiab] | 29 | the productive query for the reader literature |
MLLT1 AND (phase separation OR condensate OR droplet) | 1 | a review |
(ENL OR MLLT1) AND condensate | 12 | mostly wild-type or hotspot-mutant work |
(ENL OR MLLT1) AND condensate AND phospho* | 0 | clean translation, real zero |
(MLL-ENL OR KMT2A-MLLT1 OR KMT2A::MLLT1) AND (condensate OR phase separation OR droplet) | 1 | misses the only primary paper |
MLLT1 AND (phosphorylation OR phosphosite) | 5 | |
MLL-ENL AND (nuclear OR subnuclear OR speckle OR foci OR immunofluorescence) | 28 | |
pinometostat OR EPZ-5676 | 63 | |
(MLLT1 OR ENL OR LTG19 OR YEATS1) AND infant AND (event-free survival OR overall survival) | 22 | retired symbols included |
t(11;19) AND infant AND (event-free survival OR relapse) | 0 | parser trap, see below |
Interfant-06 AND infant AND leukemia | 19 | |
infant AND KMT2A AND (fusion partner OR translocation partner) AND (prognostic OR outcome OR survival) | 44 |
Trap 1: a t(11;19) query rewritten to a UID lookup
t(11;19) AND infant AND (event-free survival OR relapse) returned 0 records. The query translation reads "t"[All Fields] AND 11,19[UID] AND (...) 1. PubMed parsed the karyotype parentheses as grouping syntax, split (11;19) off and asked for the records whose UIDs are 11 and 19. The result is a well-formed zero that is indistinguishable from a real absence.
The same query with the band written as bare tokens, or with the gene symbol instead, returns records. This is the parenthesised-karyotype trap this project already documents, encountered live, and it is recorded here because a counted-absence claim about t(11;19) outcome built on that query would have read perfectly and been worthless.
Trap 2: the fusion-name query that misses the only primary paper
(MLL-ENL OR KMT2A-MLLT1 OR KMT2A::MLLT1) AND (condensate OR phase separation OR droplet) returns one record, and it is a review 1. The only primary paper containing a fusion-context condensate experiment does not appear, because its abstract phrases the result as 'the fusion of ENL with MLL' and the token MLL-ENL occurs only in the body text and figure legends 2.
Generalised: for a fusion, the mechanism literature is often indexed under the partner symbol alone, because the mechanistic claim is about the partner protein and the fusion is mentioned as an application. Search the partner with the mechanism term, not only the hyphenated fusion name. This is the same shape as the complex-name trap already in the standing brief.
Trap 3: a PMC HTML page that is a stub
For PMID 17855633 the four-route test behaved as follows: Europe PMC gave PMC2234781; fullTextXML returned HTTP 500; efetch db=pmc returned 8,391 characters with no <body> element; and the PMC HTML page returned HTTP 200 with 21,303 characters, which after tag stripping left 165 characters of text 1.
Route 4 is the route that has saved beats in this project, and here it returned a success-shaped failure: a page of navigation chrome and metadata with no article behind it. Measure the stripped text length, not the HTML length. A 21 kB HTML page containing 165 characters of prose passes any check that looks at the response size.
Three other on-topic papers were unreadable after all four routes plus the DOI landing page (see the register below).
Read, on topic, and deliberately not written up this cycle
Each of these would reasonably be expected in a KMT2A::MLLT1 mechanism sweep. Reason for exclusion, and what would change the decision, in one line each. Identifiers are in data.rejected and were fetched, not recalled.
- PMID 31405949, MLL-ENL leukaemia-initiation efficiency peaks neonatally in transgenic mice, and Lin28b suppresses rather than accelerates it 3. Not written up because the cell-of-origin and developmental-window beats own this material and the graph already holds
hyp-second-hit-is-developmental-window-not-mutation. Would be written up if this beat took on the question of whether reciprocal status interacts with developmental stage - which it should eventually, because this transgenic model is reciprocal-negative by construction. - PMID 42121261, internal promoters in MLLT1 and MLLT3 driving YEATS-domain-devoid isoforms 4. Read carefully because a YEATS-devoid ENL species would confound every YEATS argument in this beat. It does not: the MLLT1 internal promoter is inactive in the haematopoietic lineage and in AML, active in pituitary, brain, thyroid, uterus and lung. Recorded rather than written up because the finding's value here is a negative control.
- PMID 33144337, myeloid transformation by MLL-ENL depends strictly on C/EBP. On topic and not covered; belongs in a transcription-factor-dependency beat rather than a reciprocal-allele one.
- PMID 32766858, exploitable metabolic dependencies in MLL-ENL-induced leukaemia. A therapeutic-dependency finding orthogonal to this beat's question.
- PMID 16230385 and PMID 19029444, continuous MLL-ENL expression required for the Hox code, and MLL-ENL AML cured by oncogene ablation. Both bear on oncogene dependence rather than on partner mechanism; both would matter to a degrader dossier.
- PMID 35763909, MLLT1 SEC subunit in oesophageal cancer natural killer cells. Off-disease; retrieved by the SEC query and excluded on title.
- PMID 35395864, an ENL-selective PROTAC. Excluded because the wild-type-ENL dispensability result in t(11;19) makes ENL degradation a weak proposition in this fusion specifically; it belongs to the ENL target dossier, which exists.
On topic and unreadable after all four routes
Recorded as known-but-unreadable, with what the title asserts and nothing more.
- PMID 31283407 - Interfant-06 primary results. No PMCID in Europe PMC; the DOI resolves to an ASCO page returning HTTP 403. The abstract is available and is what this beat cites. Whether a partner-stratified outcome table exists in the full text or supplement is unknown, and no claim here depends on it being absent.
- PMID 15856011 - Zeisig 2005, ENL connects nuclear MLL fusion partners with chromatin. No PMCID; all four routes failed. Cited from abstract only; the residue-level mapping this beat needed came from a later paper instead.
- PMID 27050521 - Leukemogenic MLL-ENL fusions induce alternative chromatin states to drive a functionally dichotomous group of target genes. No PMCID; all four routes failed. The title asserts a dichotomous target-gene classification that would be directly relevant to the YEATS-dependent and YEATS-independent target split found in the murine model, and this beat could not read it.
- PMID 17855633 - Mueller 2007, the EAP purification. Readable as abstract only for the reason in Trap 3. This is the single most-cited mechanistic claim in this beat's DOT1L entry and it rests on an abstract.
| script | manual sweep, PubMed E-utilities + Europe PMC + NCBI efetch + PMC HTML |
|---|---|
| status | run |
| date | 2026-09-20 |
| result summary | 14 queries executed with translations inspected; 7 entries written; 7 on-topic papers read and deliberately not written up; 4 on-topic papers unreadable after all four routes; 3 retrieval traps encountered live. |
queries — the numbers behind the chart
[
{
"q": "MLLT1 AND DOT1L",
"n": 5
},
{
"q": "(MLL-ENL OR MLL::ENL OR KMT2A-MLLT1) AND (H3K79 OR DOT1L)",
"n": 9
},
{
"q": "MLLT1 AND (super elongation complex OR P-TEFb OR CDK9)",
"n": 9
},
{
"q": "ENL[ti] AND YEATS[tiab]",
"n": 29
},
{
"q": "MLLT1 AND (phase separation OR condensate OR droplet)",
"n": 1
},
{
"q": "(ENL OR MLLT1) AND condensate",
"n": 12
},
{
"q": "(ENL OR MLLT1) AND condensate AND phospho*",
"n": 0
},
{
"q": "(MLL-ENL OR KMT2A-MLLT1 OR KMT2A::MLLT1) AND (condensate OR phase separation OR droplet)",
"n": 1
},
{
"q": "MLLT1 AND (phosphorylation OR phosphosite)",
"n": 5
},
{
"q": "MLL-ENL AND (nuclear OR subnuclear OR speckle OR foci OR immunofluorescence)",
"n": 28
},
{
"q": "pinometostat OR EPZ-5676",
"n": 63
},
{
"q": "(MLLT1 OR ENL OR LTG19 OR YEATS1) AND infant AND (event-free survival OR overall survival)",
"n": 22
},
{
"q": "t(11;19) AND infant AND (event-free survival OR relapse)",
"n": 0,
"trap": "rewritten to 11,19[UID]"
},
{
"q": "Interfant-06 AND infant AND leukemia",
"n": 19
}
]rejected
[
{
"pmid": "31405949",
"title": "The efficiency of murine MLL-ENL-driven leukemia initiation changes with age and peaks during neonatal development.",
"reason": "developmental-window beat owns it; transgenic model is reciprocal-negative by construction",
"would_revisit_if": "this beat takes on reciprocal status x developmental stage"
},
{
"pmid": "42121261",
"title": "Tissue- and tumor-type-specific expression of internal-promoter-driven YEATS-domain-devoid isoforms of MLLT1 and MLLT3.",
"reason": "checked as a possible confounder for every YEATS argument here; the MLLT1 internal promoter is inactive in haematopoiesis and AML, so it is a negative control",
"would_revisit_if": "a leukaemia transcriptome panel shows the MLLT1 internal promoter active in any t(11;19) sample"
},
{
"pmid": "33144337",
"title": "Myeloid transformation by MLL-ENL depends strictly on C/EBP.",
"reason": "transcription-factor dependency, orthogonal to reciprocal-allele status",
"would_revisit_if": "a C/EBP dependency were shown to differ by breakpoint"
},
{
"pmid": "32766858",
"title": "Exploitable metabolic dependencies in MLL-ENL-induced leukemia.",
"reason": "therapeutic dependency beat",
"would_revisit_if": "a metabolic dependency were partner- or breakpoint-stratified"
},
{
"pmid": "16230385",
"title": "Continuous MLL-ENL expression is necessary to establish a Hox Code and maintain immortalization of hematopoietic progenitor cells.",
"reason": "oncogene-dependence rather than partner mechanism",
"would_revisit_if": "building a degrader dossier for this fusion"
},
{
"pmid": "19029444",
"title": "Acute myeloid leukemia induced by MLL-ENL is cured by oncogene ablation despite acquisition of complex genetic abnormalities.",
"reason": "oncogene-dependence rather than partner mechanism",
"would_revisit_if": "building a degrader dossier for this fusion"
},
{
"pmid": "35763909",
"title": "Down-regulation of MLLT1 super elongation complex subunit impairs the anti-tumor activity of natural killer cells in esophageal cancer.",
"reason": "off-disease; retrieved by the SEC query and excluded on title",
"would_revisit_if": "never, for this beat"
},
{
"pmid": "35395864",
"title": "A proteolysis-targeting chimera molecule selectively degrades ENL and inhibits malignant gene expression and tumor growth.",
"reason": "belongs to the ENL target dossier; wild-type ENL is dispensable in t(11;19) cells, which weakens the rationale in this fusion specifically",
"would_revisit_if": "a degrader were shown active in a t(11;19) line"
}
]unreadable after four routes — the numbers behind the chart
[
{
"pmid": "31283407",
"what_the_title_asserts": "Interfant-06 primary results in 651 infants",
"routes": "no PMCID; DOI landing page HTTP 403",
"title": "Outcome of Infants Younger Than 1 Year With Acute Lymphoblastic Leukemia Treated With the Interfant-06 Protocol: Results From an International Phase III Randomized Study.",
"authors": "Pieters R, De Lorenzo P, Ancliffe P, et al.",
"venue": "J Clin Oncol",
"year": 2019,
"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."
},
{
"pmid": "15856011",
"what_the_title_asserts": "ENL connects nuclear KMT2A fusion partners with chromatin",
"routes": "no PMCID; all four failed",
"title": "The eleven-nineteen-leukemia protein ENL connects nuclear MLL fusion partners with chromatin.",
"authors": "Zeisig DT, Bittner CB, Zeisig BB, et al.",
"venue": "Oncogene",
"year": 2005,
"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."
},
{
"pmid": "27050521",
"what_the_title_asserts": "MLL-ENL fusions induce alternative chromatin states driving a functionally dichotomous target-gene group",
"routes": "no PMCID; all four failed",
"title": "Leukemogenic MLL-ENL Fusions Induce Alternative Chromatin States to Drive a Functionally Dichotomous Group of Target Genes.",
"authors": "Garcia-Cuellar MP, Büttner C, Bartenhagen C, et al.",
"venue": "Cell Rep",
"year": 2016,
"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."
},
{
"pmid": "17855633",
"what_the_title_asserts": "ENL-associated proteins include P-TEFb and DOT1L and have H3K79 methylase and CTD kinase activity",
"routes": "fullTextXML HTTP 500; efetch db=pmc body absent; PMC HTML 21,303 chars stripping to 165",
"title": "A role for the MLL fusion partner ENL in transcriptional elongation and chromatin modification.",
"authors": "Mueller D, Bach C, Zeisig D, et al.",
"venue": "Blood",
"year": 2007,
"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."
}
]traps
[ "parenthesised karyotype rewritten to a UID lookup, returning a clean 0", "fusion-name query misses the only primary fusion-context paper", "PMC HTML route 4 returns a 21 kB stub containing 165 characters of text" ]
- An enumeration entry can create false confidence: recording what was rejected makes a sweep look complete when it is only documented. Nothing here establishes that the queries were the right queries.
- Two of the three traps documented here were found by accident rather than by a check that would catch them next time. A stripped-text length assertion and a querytranslation assertion belong in tooling, not in a prose entry.
- Can PMID 27050521 be reached by inter-library loan or by an author request? Its title asserts a dichotomous MLL-ENL target-gene classification that would test the YEATS-dependent and YEATS-independent target split this beat found in a murine model.
- Does a partner-stratified outcome table exist in the Interfant-06 full text or supplement? Nothing in this beat depends on it being absent, and one reader with journal access could close the question in a minute.
- Should the whole graph's counted-absence claims be re-run with a stripped-text length check on route 4? At least one of them may have been recorded against a stub.
- How many of this graph's counted absences used a parenthesised karyotype? That query form returns a clean zero for every band, and the failure is invisible in the result.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
The queries listed in this entry were executed against PubMed E-utilities on 2026-09-20 with the returned query translations inspected; `(ENL OR MLLT1) AND condensate AND phospho*` returns 0 with a clean translation, and `t(11;19) AND infant AND (event-free survival OR relapse)` returns 0 with the translation `"t"[All Fields] AND 11,19[UID] AND (...)`.
PubMed E-utilities esearch, db=pubmed, batch run 2026-09-20 with querytranslation inspected for every query · NCBI Entrez Programming Utilities · NCBI E-utilities 2026What this evidence cannot show
Counts are as of one day and will drift. Symbol coverage varies by query: the outcome queries included the retired symbols LTG19 and YEATS1, the mechanism queries did not. `ENL` as a bare token is contaminated by erythema nodosum leprosum and by ordinary English, so non-zero counts from queries containing it are upper bounds and were title-checked before use.
Evidence that argues against this (1)
- A counted absence is only as good as its symbol set; this project has previously under-counted a partner literature tenfold by searching current HGNC symbols only, so the zeros here should be re-run with alias expansion before any of them is treated as settled. Meyer C et al. 2023
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The only primary paper containing a fusion-context condensate experiment for this partner phrases the result in its abstract as 'the fusion of ENL with MLL can boost SEC phase separation', so the token MLL-ENL appears only in its body and figure legends.
ENL initiates multivalent phase separation of the super elongation complex (SEC) in controlling rapid transcriptional activation. · Guo C, Che Z, Yue J, et al. · Science Advances 2020What this evidence cannot show
This is an observation about one paper's phrasing and PubMed's field indexing on one date. PubMed indexes titles and abstracts, not full text, so the generalisation holds only for abstract-level retrieval; a full-text search service would find it.
Evidence that argues against this (1)
- Searching the partner symbol alone with the mechanism term does retrieve this paper, so the trap is specific to fusion-name queries rather than a general failure of PubMed retrieval. NCBI Entrez Programming Utilities 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In transgenic mice, MLL-ENL induced in fetal or neonatal life caused AML in almost all animals while induction in adults left most surviving beyond a year despite sustained transgene expression; initiation was most efficient in neonates, transient suppression in neonates prevented AML in most mice, and the fetal regulator Lin28b suppressed rather than accelerated initiation.
The efficiency of murine MLL-ENL-driven leukemia initiation changes with age and peaks during neonatal development. · Okeyo-Owuor T, Li Y, Patel RM, et al. · Blood Advances 2019What this evidence cannot show
A transgenic murine model expressing MLL-ENL from a transgene, therefore reciprocal-negative by construction and not a model of the reciprocal-positive half of the human population. Read from abstract only. Transplantation stress mitigated the developmental barrier, which means the readout is sensitive to the assay used to score it.
Evidence that argues against this (1)
- Because the model is single-allele, it cannot address whether the reciprocal product modifies the developmental window, and this graph already records that mouse models of this fusion differ systematically in reciprocal status depending on how they were built. Hu H et al. 2023
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
An internal promoter in the MLLT1 locus drives a YEATS-domain-devoid ENL isoform, and that promoter is inactive in the haematopoietic lineage and in AML while being functional in pituitary gland, brain, thyroid, uterus and lung; the analogous MLLT3 internal promoter is by contrast active in healthy haematopoietic stem cells and in poor-prognosis AML blasts.
Tissue- and tumor-type-specific expression of internal-promoter-driven YEATS-domain-devoid isoforms of MLLT1 and MLLT3. · Dailhau S, Zamani A, Pfeifer R, et al. · Biomarker Research 2026What this evidence cannot show
Read from abstract only. Promoter activity is inferred from transcript expression across tissue panels rather than from a direct promoter assay in leukaemic cells, and 'inactive in the haematopoietic lineage and AML' is a statement about the datasets examined. Relevant to this beat as a negative control rather than as a positive finding.
Evidence that argues against this (1)
- The same work finds the MLLT3 internal promoter active in haematopoietic stem cells and in poor-prognosis AML, so the haematopoietic silence of the MLLT1 internal promoter is a paralogue-specific claim that a larger leukaemia transcriptome panel could overturn. Dailhau S et al. 2026
- related to →KMT2A::MLLT1 (MLL-ENL), t(11;19)(q23;p13.3)
audit trail for the mechanism, epigenetics, localisation and outcome sweep of this fusion
four on-topic papers unreadable after all four routes, with what each title asserts and nothing more
records that the EAP claim underpinning this entry rests on an abstract, because route 4 returned a stub
checks and clears the YEATS-devoid internal-promoter isoform as a confounder for every YEATS argument in this beat
adds a transgenic MLL-ENL model that is reciprocal-negative by construction, with a neonatal initiation peak
records Interfant-06 as unreadable and states that no claim in the outcome entry depends on a partner table being absent from it
the retrieval trap that explains why the fusion-context condensate literature looks empty and is not
MLLT1 internal promoter drives a YEATS-devoid isoform that is inactive in haematopoiesis and AML