AnalysisstrongQ36-menin-interactome

Menin's 33 curated partners, enumerated: 10 written, 10 discarded as mutation-directed or single-method, 3 that are not proteins, and what was read and rejected

also called menin interactome enumeration · MEN1 partner ledger · Q36 bounded rejections

This is the ledger for the whole beat: every protein a database says menin binds, what the evidence for each actually is, which ones got written up, which were rejected and why. Recording the rejections matters as much as the entries — otherwise the next person repeats the same search and cannot tell what was already considered and dismissed.

The same thing, for a scientist

A full MITAB 2.7 pull of IntAct on MEN1 (O00255) on 2026-09-20 returns **59 rows covering 33 distinct partners** [E1]. This beat's disposition of them: **10 partners already had entries or were written this cycle**; **10 were rejected as mutation-directed neo-interactions or single-method BRET rows from one screen** (`ana-menin-neoppi-trap-ten-intact-partners-from-one-mutation-screen`); **3 are not proteins at all** but menin target genes curated from ChIP data; and **the remainder are proximity-labelling or large-scale rows recorded here as bounded rejections rather than written up**. The headline number in the dispatch brief — *4 of 30 MEN1 partners have an entry, 13.3%, worst of six baits* — is now **14 of 33, 42.4%**, and, more usefully, **the fraction of partners whose evidence is misread if the method column is ignored is 13 of 33 (39.4%)**. The second number is the one that should change how this project reads IntAct. **This entry also records what was read and NOT used**, which the brief identifies as the half of a negative this graph is blind to: eleven papers and four candidate partners were examined and dropped, each with the reason and what would change it.

Go deeper

Full disposition of the 33 partners

partner	source	method	disposition
KMT2A	15199122, 18598942, 20153263	anti-bait co-IP x3	already written
PSIP1	18598942	anti-bait co-IP	already written
ESR1	31527615	TAP	already written (this cycle, another queen)
HCFC1	— (not in this pull; via HBM literature)	—	already written
CSNK2A1	— (paralogue route)	—	already written
FANCD2	12874027	co-IP x3 + MS	written: `gene-fancd2`, `interface-menin-fancd2`
NFKB1	11526476	co-IP + pull-down	written: `gene-nfkb1`
NFKB2	11526476	pull-down	written: `gene-nfkb2`
RELA	11526476	pull-down	written: `gene-rela`
MYH9	14508515	co-IP, cosedimentation, colocalisation	written: `gene-myh9`
CSNK2A2	27705803	TAP	written: `gene-csnk2a2`
HCFC2	15199122	molecular sieving	written: `gene-hcfc2`
EBF3	17353931	anti-bait co-IP (screen)	written: `gene-ebf3`
BRAF, AKT1, TP53, EGFR, PTEN	35512704	BRET, `mutation causing an interaction`	rejected — neo-interactions
SPOP, SMAD4, SMARCA4, FBXW7, PTPN11	35512704	BRET, `mutation with no effect`	rejected — single-method over-expression
HOXA7, HOXA9, MEIS1	18598942	chromatin immunoprecipitation	not partners — target genes
RPRM	17353931	anti-bait co-IP (screen)	bounded rejection, below
CD226, ZC3HC1	33961781	anti-tag co-IP (BioPlex)	bounded rejection, below
AFG2A	38554706	TAP	bounded rejection, below
BUD13, NPM1, RPS11	39251607	BioID	bounded rejection, below
H3C1	29568061	BioID	bounded rejection, below
WDR5	27705803	TAP	already written elsewhere

Bounded rejections — read, considered, not written

Recorded per the standing rule, with the identifier, the reason, and what would change the decision. These are not absences; they are decisions.

RPRM (Reprimo), PMID 17353931. Single row in the same 338-bait screen as EBF3, no orthogonal confirmation in nineteen years, no leukaemia literature, and no mechanistic story connecting a p53-responsive cell-cycle gene to menin. EBF3 was written and RPRM was not, on one ground only: EBF3's family determines the lineage of the index disease and Reprimo's does not. Would change it: any directed co-IP, or evidence that RPRM is expressed in KMT2A-rearranged blasts and menin-dependent.

CD226 and ZC3HC1, PMID 33961781 (BioPlex). Anti-tag co-IP rows from a proteome-scale network in HEK293T and HCT116. CD226 (DNAM-1) is an NK and T-cell surface receptor — its presence in a HEK293T over-expression network is not evidence of a nuclear menin complex. ZC3HC1 (NIPA) is an SCF substrate-recognition subunit with a plausible nuclear location and no menin literature. Would change it: recovery in a haematopoietic-cell interactome, or a directed experiment.

AFG2A, PMID 38554706. Recovered in a TAP from a study of the SPATA5-SPATA5L1 ATPase complex directing replisome proteostasis. Interesting on its face given this beat's replication-stress neighbourhood, and rejected because it is one row in a network whose bait was not menin and whose subject is a different complex. Would change it: a menin-bait experiment recovering it, or evidence that menin is a replisome-proteostasis client.

BUD13, NPM1 and RPS11, PMID 39251607. BioID rows from a systematic post-transcriptional regulatory module study. Proximity labelling is not binding — a BioID hit means within roughly 10 nm for long enough to be biotinylated. NPM1 is a nucleolar abundant protein and the single most common BioID background protein in any nuclear bait; RPS11 is a ribosomal protein, same objection. BUD13 is a retention-and-splicing complex subunit and is the only one of the three with a non-trivial reading. Would change it: a menin co-IP with BUD13, or a functional splicing phenotype under menin loss. The graph's mechanism-rna-layer-and-menin-inhibition is where such a result would attach.

H3C1, PMID 29568061. A BioID row for a histone H3 gene product from a MAC-tag methods paper. Menin being proximal to histone H3 is expected of any chromatin protein and carries no information. Would change it: nothing short of a specific modified-histone binding result.

Papers read and judged off-beat. PMID 36909940 (menin represses gastric cancer proliferation via IQGAP1) — a real directed menin interaction, rejected because IQGAP1 is not in menin's IntAct record, the system is gastric cancer, and it would have been a fifth tissue with no leukaemia bridge. PMID 35509630 (menin deletion in mouse mesenchymal stem cells), 30220511 (menin deficiency and depressive-like behaviour via astrocyte neuroinflammation), 34971677 (habenular inflammatory cytokines), 16322332 (NF-kB in human parathyroid tumours), 15691381 (global gene expression in MEN1 neuroendocrine tumours), 22275377 (menin in-frame deletion mutant and TGF-beta), 22016823 (menin represses tumorigenesis, review-like), 12755956 (functional studies of the MEN1 gene, 2003 review), 12114442 (genistein and NF-kB in PC3 cells — returned by a menin/NF-kB query and containing no menin), 42697596, 35681273, 31428572, 30942098, 26691941 (five cancer-genomics papers returned by the menin/Fanconi query, all incidental co-occurrence of MEN1 in a gene panel). None is cited anywhere in this beat.

A laboratory judged inapplicable. The Hua laboratory (University of Pennsylvania) is the source of the founding menin-FANCD2 paper and publishes continuously on menin in endocrine and epithelial systems. Nothing from that body of work after 2003 addresses the FANCD2 link, and none of it is haematopoietic. Would change it: any Hua-laboratory paper returning to menin and DNA repair, which a standing alert on menin AND (DNA repair OR DNA damage) would catch.

What this changes about reading IntAct

Three mechanical rules, each earned by a specific row in this pull.

Read the method column before the interactor column. Thirteen of 33 partners (39.4%) are misread without it: 10 BRET rows from a mutation screen and 3 ChIP-curated target genes.
Read the feature columns. mutation causing an interaction versus mutation with no effect is the entire difference between a false partner and a weak one, and it lives in MITAB columns 36 and 37, which most parsers drop.
Count source publications, not rows. Menin's 59 rows come from 13 publications, and 31 rows come from one. A partner with four rows from one paper is not better supported than a partner with one row from one paper; it has been assayed more times by the same people.

The brief's framing — 4 of 30 partners have an entry, 13.3%, the worst of six baits — was a fair measurement and a good reason to dispatch this beat. But the coverage gap was smaller than it looked, because a third of the partner list should not be written at all.

Facts and figures

script	manual MITAB 2.7 pull and parse; no committed script
status	run
result summary	33 distinct MEN1 partners from 13 publications across 59 rows. 14 now have entries (42.4%, up from the brief's 13.3%). 13 of 33 (39.4%) are misread if the method and feature columns are ignored: 10 mutation-screen BRET rows and 3 ChIP-curated target genes. Eleven papers and four candidate partners were read and rejected, each recorded with its reason.

coverage

bounded rejections

inputs

[
  "https://www.ebi.ac.uk/Tools/webservices/psicquic/intact/webservices/current/search/interactor/O00255?format=tab27"
]

coverage — the numbers behind the chart

{
  "partners_total": 33,
  "partners_with_entries_after_this_cycle": 14,
  "coverage_pct": 42.4,
  "brief_stated_coverage_pct": 13.3,
  "partners_that_should_not_be_written": 13,
  "source_publications": 13,
  "rows_from_single_publication_35512704": 31
}

written this cycle

[
  "gene-fancd2",
  "gene-nfkb1",
  "gene-nfkb2",
  "gene-rela",
  "gene-myh9",
  "gene-hcfc2",
  "gene-ebf3"
]

bounded rejections — the numbers behind the chart

[
  {
    "item": "RPRM",
    "pmid": "17353931",
    "reason": "single row in a 338-bait screen, no orthogonal confirmation, no leukaemia literature, no mechanistic route",
    "would_change": "a directed co-IP, or expression plus menin-dependence in KMT2A-rearranged blasts",
    "title": "Large-scale mapping of human protein-protein interactions by mass spectrometry.",
    "authors": "Ewing RM, Chu P, Elisma F, et al.",
    "venue": "Mol Syst Biol",
    "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."
  },
  {
    "item": "CD226",
    "pmid": "33961781",
    "reason": "BioPlex anti-tag co-IP in HEK293T/HCT116; CD226 is an NK/T surface receptor with no nuclear role",
    "would_change": "recovery in a haematopoietic interactome",
    "title": "Dual proteome-scale networks reveal cell-specific remodeling of the human interactome.",
    "authors": "Huttlin EL, Bruckner RJ, Navarrete-Perea J, et al.",
    "venue": "Cell",
    "year": 2021,
    "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."
  },
  {
    "item": "ZC3HC1",
    "pmid": "33961781",
    "reason": "BioPlex row only; plausible nuclear location but no menin literature and no mechanism",
    "would_change": "a directed experiment or a second independent map",
    "title": "Dual proteome-scale networks reveal cell-specific remodeling of the human interactome.",
    "authors": "Huttlin EL, Bruckner RJ, Navarrete-Perea J, et al.",
    "venue": "Cell",
    "year": 2021,
    "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."
  },
  {
    "item": "AFG2A",
    "pmid": "38554706",
    "reason": "one TAP row from a study whose bait and subject were the SPATA5-SPATA5L1 replisome proteostasis complex, not menin",
    "would_change": "recovery with menin as bait, or evidence menin is a replisome-proteostasis client",
    "title": "The SPATA5-SPATA5L1 ATPase complex directs replisome proteostasis to ensure genome integrity.",
    "authors": "Krishnamoorthy V, Foglizzo M, Dilley RL, et al.",
    "venue": "Cell",
    "year": 2024,
    "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."
  },
  {
    "item": "BUD13, NPM1, RPS11",
    "pmid": "39251607",
    "reason": "BioID proximity, not binding; NPM1 and ribosomal proteins are canonical nuclear BioID background",
    "would_change": "a menin co-IP with BUD13, or a splicing phenotype under menin loss",
    "title": "Systematic identification of post-transcriptional regulatory modules.",
    "authors": "Khoroshkin M, Buyan A, Dodel M, et al.",
    "venue": "Nat Commun",
    "year": 2024,
    "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."
  },
  {
    "item": "H3C1",
    "pmid": "29568061",
    "reason": "BioID proximity to histone H3 from a methods paper; expected of any chromatin protein and uninformative",
    "would_change": "a specific modified-histone binding result",
    "title": "An AP-MS- and BioID-compatible MAC-tag enables comprehensive mapping of protein interactions and subcellular localizations.",
    "authors": "Liu X, Salokas K, Tamene F, et al.",
    "venue": "Nat Commun",
    "year": 2018,
    "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."
  },
  {
    "item": "IQGAP1 (menin partner in gastric cancer)",
    "pmid": "36909940",
    "reason": "a real directed menin interaction, but absent from menin's IntAct record, in gastric cancer, with no bridge to leukaemia; would have been a fifth tissue",
    "would_change": "any haematopoietic replication",
    "title": "Menin represses the proliferation of gastric cancer cells by interacting with IQGAP1.",
    "authors": "Ren F, Guo Q, Zhou H",
    "venue": "Biomed Rep",
    "year": 2023,
    "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."
  },
  {
    "item": "eleven further papers returned by the beat's queries",
    "reason": "off-beat on inspection: mesenchymal stem cells, murine behaviour, parathyroid tumours, neuroendocrine expression profiling, TGF-beta, reviews, and cancer-genomics panels containing MEN1 incidentally",
    "would_change": "nothing foreseeable; recorded so the searches are not repeated blindly",
    "pmids": [
      "35509630",
      "30220511",
      "34971677",
      "16322332",
      "15691381",
      "22275377",
      "22016823",
      "12755956",
      "12114442",
      "42697596",
      "26691941"
    ],
    "pmids_verified": "All 11 resolve on PubMed esummary, checked 2026-09-20, and each matches the rejection reason recorded here. Previously these were packed into a single 'pmid' string, where verify-citations.py could not read them: a comma-joined value is not a PMID, so eleven identifiers sat unchecked inside a field whose whole purpose is to be checked."
  }
]

reading rules derived

[
  "read the method column before the interactor column",
  "read MITAB feature columns 36 and 37 for mutation annotations",
  "count source publications, not rows"
]

deduplicated with concurrent beat

{
  "id": "gene-csnk2a2",
  "note": "A gene-csnk2a2 entry was drafted in this beat and deleted on discovering that the concurrently running P-TEFb reservoir beat had already written data/graph/nodes/ptefb-reservoir/gene-csnk2a2.json, which is stronger on the dependency and expression side. Two menin-specific findings from the deleted draft are NOT in that entry and should be merged into it by the coordinator: (1) IntAct records a MEN1-CSNK2A2 association by tandem affinity purification from PMID 27705803, the human Polycomb complexome map, in which menin, CSNK2A2 and WDR5 co-purify - this is the only menin-CK2 evidence that exists, and it is an AP-MS co-purification, which is weak because CK2 subunits are among the commonest recoveries in nuclear AP-MS; (2) CX-4945 is ATP-competitive and the two CK2 catalytic subunits share their ATP site, so every CX-4945 phenotype cited in gene-csnk2a1 - including the MLL-AF9 leukaemia-stem-cell depletion - is a CK2-level result and cannot be attributed to CSNK2A1 specifically."
}

How this could bite us back

The main risk this entry guards against is not a wrong fact but a wrong denominator: a coverage metric that counts database rows will keep dispatching beats to write partners that should not be written.
Bounded rejections are only useful if they are revisited. Each one here names what would change it, and none will be checked again unless somebody schedules it.

Still open

How many of the other five core scaffold baits in the coverage analysis have partner lists contaminated by PMID 35512704 or by ChIP-curated target genes? The same parse should be run on all six.
Should the project add a standing IntAct parser that filters on method and feature columns, so this class of error cannot recur?
Is IntAct the right source at all? BioGRID and STRING carry partners IntAct does not, and 33 is a floor rather than a count.
Of the eight partners written this cycle, which single one would most change the therapeutic picture if a directed experiment confirmed it? On present evidence RELA, and the experiment is an in vitro competition with revumenib.

Every source behind this entry (4)

E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
A MITAB 2.7 PSICQUIC pull of IntAct on interactor O00255 (MEN1) on 2026-09-20 returns 59 rows covering 33 distinct partners from 13 source publications, of which 31 rows and 10 partners derive from PMID 35512704 and 3 'partners' (HOXA7, HOXA9, MEIS1) are curated from chromatin immunoprecipitation and are target genes rather than proteins.
IntAct molecular interaction database, PSICQUIC MITAB 2.7 query on interactor O00255 (MEN1), retrieved 2026-09-20 · IntAct (EMBL-EBI) 2026
What this evidence cannot show
PSICQUIC is one view of IntAct and does not necessarily match the aggregated web interface: the dispatch brief listed four source PMIDs for CSNK2A2 where this pull returns one, and listed 30 partners where this returns 33. The counts describe this response on this date. IntAct is also not the whole curated interactome - BioGRID, STRING and the primary literature each contain partners IntAct does not, so 33 is a floor.
Evidence that argues against this (1)
- The brief's own count of 30 partners came from a different IntAct view, and the discrepancy is a reminder that a partner count is a property of a query rather than of the protein. Neither number should be quoted without the query that produced it. Oncofusion knowledge graph, analyses/core-complex-interactome-gap 2026
E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Ten of menin's 33 IntAct partners derive from a single BRET screen designed to find interactions created by cancer mutations; five carry the IntAct feature 'mutation causing an interaction' with the alleles BRAF V600E, AKT1 E17K, TP53 R282W, EGFR residue 858 and PTEN residue 173, and five carry 'mutation with no effect'.
Systematic discovery of mutation-directed neo-protein-protein interactions in cancer. · Mo X, Niu Q, Ivanov AA, Tsang YH, Tang C, Shu C, Li Q, Qian K, Wahafu A, Doyle SP, Cicka D, Yang X, Fan D, Reyna MA, Cooper LAD, Moreno CS, Zhou W, Owonikoko TK, Lonial S, Khuri FR, Du Y, Ramalingam SS, Mills GB, Fu H · Cell 2022
PMID 35512704 DOI
What this evidence cannot show
The allele assignments come from IntAct's curation resolved against ClinVar and dbSNP, not from the paper's own text: MEN1 appears in the paper's body only in the key resources table as an antibody, and all menin results are in supplementary tables that were not parsed. Full detail is in `ana-menin-neoppi-trap-ten-intact-partners-from-one-mutation-screen`.
Evidence that argues against this (1)
- A curated interaction inherits every limitation of the single experiment behind it. Anti-bait co-immunoprecipitation and affinity-capture mass spectrometry cannot distinguish a direct contact from a shared-complex association, tagged-bait overexpression can create contacts that do not occur at endogenous stoichiometry, and a partner curated from one publication has by definition never been independently replicated. The enumeration is accurate about the database; the database is not a statement about the blast. Heppner C et al. 2001
E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The Ewing large-scale mass-spectrometry map, which is the sole source for the menin partners EBF3 and RPRM, screened 338 FLAG-tagged baits and reported 24,540 potential interactions filtered to 6,463 by empirical criteria and a calculated confidence score, so an individual row carries the screen's aggregate false-discovery properties rather than the confidence of a directed experiment.
Large-scale mapping of human protein-protein interactions by mass spectrometry. · Ewing RM, Chu P, Elisma F, Li H, Taylor P, Climie S, McBroom-Cerajewski L, Robinson MD, O'Connor L, Li M, et al. · Mol Syst Biol 2007
PMID 17353931 DOI
What this evidence cannot show
This is the standard caveat on any large-scale screen and is not a criticism of the dataset, which was the first of its kind and is widely used. It bears on how a single row should be graded, not on whether the study was sound.
Evidence that argues against this (1)
- The filter IS the result. 24,540 potential interactions reduced to 6,463 means a threshold was chosen, and partners recovered from a single bait near that threshold - which is what EBF3 and RPRM are - would appear or vanish under a different cut. A large-scale map's sensitivity and its false-positive rate are the same dial. Heppner C et al. 2001
E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Three of menin's IntAct partners - BUD13, NPM1 and RPS11 - come from proximity-dependent biotin identification, which reports spatial proximity within roughly 10 nm rather than binding; NPM1 and ribosomal proteins are among the commonest background recoveries in nuclear BioID experiments.
Systematic identification of post-transcriptional regulatory modules. · Khoroshkin M, Buyan A, Dodel M, Navickas A, Yu J, Trejo F, Doty A, Baratam R, Zhou S, Lee SB, Joshi T, Garcia K, Choi B, Miglani S, Subramanyam V, Modi H, Corices C, Ghandehari R, Winslow MM, Mardakheh FK, Kulakovskiy IV, Goodarzi H · Nat Commun 2024
PMID 39251607 DOI
What this evidence cannot show
Cited for the method and for the fact that these rows exist, not for a claim the paper makes about menin. The characterisation of NPM1 and ribosomal proteins as BioID background is a general methodological point rather than a finding of this particular study, and this graph's own proximity-interactome beat records the same promiscuity problem for KMT2A baits.
Evidence that argues against this (1)
- This project's own proximity beat found that 25 of 50 top KMT2A proximity hits were promiscuous across unrelated baits, so a BioID row is discounted here on internal evidence as well as on general principle. Oncofusion knowledge graph, ana-kmt2a-proximity-promiscuity-25-of-50 2026

Connected to

← related to577 citations carried a bare PMID with no title, so the audit built to catch a fabricated title could not see them: the third dropped denominator in this project
HELD ELEVEN PMIDS PACKED INTO ONE FIELD, unread by every checker; all eleven verified and re-stored as an array.
supports →Ten of menin's 33 IntAct partners come from one BRET screen for MUTATION-created interactions — and five are gains caused by a named allele
the enumeration's rejection of ten partners rests on this analysis
related to →MEN1 / menin — the adaptor the fusion cannot work without
full ledger of menin's 33 curated partners and their disposition
analogous to →First systematic enumeration of the KMT2A fusion interactome, and the diff against this graph
the same enumeration pattern applied to the MEN1 bait
← part ofEBF3 — a single large-scale MS hit against menin, in a B-lineage transcription factor family, which is the only reason it is worth recording
written at weak confidence under the enumeration's single-screen-hit rule
← related toAll nine unwritten MEN1 partners trace to exactly two publications, neither of which mentions menin outside an antibody catalogue row
same partner list, re-pulled 2026-09-23 through IntAct REST rather than PSICQUIC