Nobody has ever shown 14-3-3 binding directly to KMT2A: five records across every paralogue alias, and one of them is about AF4
also called 14-3-3 KMT2A absence · YWHA KMT2A literature · bounded absence 14-3-3
KMT2A turns up as a 14-3-3 partner in two big screens. We searched the published literature using every name the seven 14-3-3 proteins go by and found no paper that ever tested the pair directly - five records in total, and the only relevant one is about KMT2A's fusion partner, not KMT2A.
The same thing, for a scientist
The dispatch asked whether there is evidence of 14-3-3 binding to KMT2A beyond the two interactome datasets. **There is not.** The alias-expanded PubMed query over all seven paralogue symbols plus the family name returns **5 records**. Title-checking all five: **one is on-beat and it is about AF4, not KMT2A** (14-3-3 theta binds AFF1 Ser588 and complexes with MLL-AF4); the other four are off-beat, two of them in ways that would fool a count. **Zero papers report a pulldown, a co-immunoprecipitation, a motif or a functional consequence of 14-3-3 binding KMT2A directly.** Combined with the phosphosite verdict on S821 and the zero-canonical-motif result on AFF1, the state of the question is: the proximity observation is real and reproduced, the mechanism is unknown, and the one demonstrated 14-3-3 contact in this system is on the partner.
The queries and the counts
Run against PubMed 2026-09-19:
| query | records |
|---|---|
(MLL OR MLL1 OR KMT2A OR ALL-1 OR HRX OR HTRX) AND (14-3-3 OR YWHAB OR YWHAE OR YWHAG OR YWHAH OR YWHAQ OR YWHAZ OR SFN OR stratifin) | 5 |
(KMT2A OR MLL1) AND YWHAG | 0 |
(KMT2A OR MLL OR MLL1) AND (14-3-3 zeta OR epsilon OR sigma OR tau OR gamma OR beta OR eta) | 0 |
MLL-AF4 AND 14-3-3 | 1 |
AFF1 AND 14-3-3 | 2 |
AF4 AND 14-3-3 | 1 |
YWHAG is the paralogue that reproduces across both interactome datasets, and (KMT2A OR MLL1) AND YWHAG returns zero records. The single most reproduced 14-3-3 relation in this system has never been written about.
Per the standing alias rule, retired and alternative symbols were included (ALL-1, HRX, HTRX for KMT2A; the Greek-letter forms and SFN/stratifin for the family). Per the precision half of the same rule, every returned record was title-checked rather than counted.
What the five records actually are
| PMID | verdict | why |
|---|---|---|
| 31493143 | on-beat, partner side | 14-3-3 theta binds AF4 Ser588, complexes with MLL-AF4 1 |
| 32792914 | rejected | MLL1 drives transcription of ASK-1 in ischaemic neurons; the 14-3-3 in the abstract is part of the ASK-1/STRAP/14-3-3 complex MLL1 regulates, not a KMT2A partner. This is the record most likely to be miscounted as a hit |
| 26800457 | rejected, alias collision | C. elegans cell-fusion genes EFF-1 and AFF-1. The nematode AFF-1 is a fusogen, unrelated to human AFF1 |
| 28094252 | rejected | LIM-protein/CHK2/CDC25/14-3-3 radiosensitivity; MLL1 appears only as a transcriptional driver of FHL1 |
| 31138789 | rejected | ZNF479/MT-1 in hepatocellular carcinoma; 14-3-3 epsilon and the MLL complex are separate parts of one signalling chain |
Four of five are rejections, and three of them would read as hits to anyone counting titles containing both terms. That is the shape the brief warns about: a confident, bounded, wrong number is easy to produce here.
What would overturn the absence
- A 14-3-3 immunoprecipitation from any cell type blotted for KMT2A, or the reverse.
- A KMT2A residue whose mutation abolishes 14-3-3 recovery in the published interactome data - that dataset exists and could be re-mined without new experiments.
- Any paper that discusses the KMT2A rows in the seven-paralogue interactome. This beat found none; the rows are curated into IntAct and, as far as these queries reach, never discussed.
- A non-English or pre-1990 report under a retired symbol. The alias set used here is broad but not exhaustive.
The absence is of published direct evidence, not of the interaction. Two independent datasets put KMT2A in 14-3-3 neighbourhoods and that remains the single best-reproduced proximity relation in the graph's KMT2A record.
| script | analyses/14-3-3-kmt2a-literature/run.py |
|---|---|
| status | run |
| result summary | Zero published direct evidence of 14-3-3 binding KMT2A. The only on-beat record concerns AFF1 S588. Four of five records are rejections and three would pass an untitle-checked count. |
queries — the numbers behind the chart
[
{
"q": "(MLL OR MLL1 OR KMT2A OR ALL-1 OR HRX OR HTRX) AND (14-3-3 OR YWHAB OR YWHAE OR YWHAG OR YWHAH OR YWHAQ OR YWHAZ OR SFN OR stratifin)",
"n": 5
},
{
"q": "(KMT2A OR MLL1) AND YWHAG",
"n": 0
},
{
"q": "(KMT2A OR MLL OR MLL1) AND (14-3-3 zeta OR 14-3-3 epsilon OR 14-3-3 sigma OR 14-3-3 tau OR 14-3-3 gamma OR 14-3-3 beta OR 14-3-3 eta)",
"n": 0
},
{
"q": "MLL-AF4 AND 14-3-3",
"n": 1
},
{
"q": "AFF1 AND 14-3-3",
"n": 2
},
{
"q": "AF4 AND 14-3-3",
"n": 1
}
]on beat
[ "31493143" ]
rejected
[ "32792914", "26800457", "28094252", "31138789" ]
- Re-mine the seven-paralogue interactome supplementary data for the KMT2A rows: which peptides, which paralogues, what scores? That is a desk exercise with no new experiment and nobody has done it.
- Is the KMT2A-14-3-3 recovery abolished by phosphatase treatment of the lysate? That one control distinguishes a phospho-dependent contact from a co-purification artefact.
- Do the four rejected papers contain unreported KMT2A/14-3-3 data in their figures? Abstract-level rejection is the weakest step in this analysis.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
An alias-expanded PubMed query covering all seven 14-3-3 paralogue symbols, the family name and four KMT2A symbols including the retired ALL-1, HRX and HTRX returns five records; title-checking all five finds one on-beat paper, which reports 14-3-3 theta binding AF4 Ser588 and forming a complex with MLL-AF4, and four off-beat records. A query pairing KMT2A or MLL1 with YWHAG, the paralogue reproduced across both interactome datasets, returns zero records.
Crosstalk between 14-3-3θ and AF4 enhances MLL-AF4 activity and promotes leukemia cell proliferation. · Fioretti T, et al. · Cell Oncol (Dordr) 2019What this evidence cannot show
Queries run 2026-09-19 against live PubMed esearch. PubMed indexing is not a census of the literature: a finding reported only in a supplementary table, a preprint, a thesis or a non-indexed journal would not appear. The alias set covers the current and three retired KMT2A symbols and all seven paralogue gene symbols plus Greek-letter forms, but not every historical spelling. The counted absence is of PubMed-indexed titles and abstracts, not of experiments.
Evidence that argues against this (1)
- KMT2A is recovered as prey by six of the seven 14-3-3 paralogues in a curated interactome and independently by YWHAG in a second proximity study. The literature being silent about a relation does not mean the relation is absent; here it means nobody has written about a result that is sitting in two public datasets. Segal D et al. 2023
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Three of the four rejected records contain both a KMT2A-family symbol and a 14-3-3 term in a way that would be counted as a hit by an untitle-checked query: in PMID 32792914 MLL1 drives transcription of ASK-1 whose complex includes 14-3-3; in PMID 28094252 MLL1 drives FHL1 transcription in a CDC25/14-3-3 pathway; in PMID 26800457 the '14-3-3' is genuine but the 'AFF-1' is a Caenorhabditis elegans fusogen unrelated to human AFF1.
Mixed Lineage Leukemia 1 Promoted Neuron Apoptosis in Ischemic Penumbra via Regulating ASK-1/TNF-α Complex. · Feng Z, et al. · Front Neuroanat 2020What this evidence cannot show
Judged from abstracts, read 2026-09-19. A paper can contain a co-immunoprecipitation not mentioned in its abstract, so these rejections are abstract-level and would be overturned by a full-text finding. The AFF-1 collision is a species-level homonym and is noted so a future enumeration does not rediscover it as a lead.
Evidence that argues against this (1)
- Rejecting a paper on its abstract is exactly the error this project has logged elsewhere: an empty or uninformative abstract field carries no information about what a paper contains, and two papers previously written off in this graph turned out to be open-access full texts with directly relevant content. These four should be re-checked at full text before the absence is treated as final. Xu X et al. 2017
bounds the literature behind the proximity claim: zero direct reports
- related to →KMT2A / MLL1 — the gene that breaks
counted absence over all seven 14-3-3 paralogue aliases
counted absence 2
gamma-specific extension of that family-level absence: YWHAG AND nuclear = 4 records, all off-beat on title-check; YWHAG AND leukemia = 3, all incidental