Oncofusion

start / graph / gene-ibtk

Geneweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q75-dot1l-partners

IBTK - the only one of the six DOT1L spokes with a reciprocal test, and its own pulldown in the same study does not find DOT1L

also called BTKI · KIAA1417 · IBtk · Inhibitor of Bruton tyrosine kinase

IBTK is a protein that dampens Bruton's tyrosine kinase, a central switch in B cells. A DOT1L pulldown found a trace of it - about one IBTK per 476 DOT1L - and when the same study pulled on IBTK instead, DOT1L was not there.

The same thing, for a scientist

IBTK (Q9P2D0, 1,353 aa, reviewed) is the substrate-receptor of a CUL3 ubiquitin ligase that inhibits BTK, and is the one of the six unwritten DOT1L partners with genuine a-priori interest in a B-lineage disease. Its DOT1L evidence is the weakest of the set on stoichiometry - **0.0021, one IBTK per 476 DOT1L**, class B, prey-bait profile correlation 0.156, `core.interactor` not set [E2] - and it is the only one with a **reciprocal arm in the same dataset**: IBTK was itself a bait cell line (MCP_ky_0004631), GFP-IBTK recovered 23 partners in the same HeLa background, and **DOT1L is not among them** [E3]. Two identifier traps sit on this row. The IntAct interactor and the accession quoted in this project's coverage recheck are **E7EPI0, an UNREVIEWED TrEMBL entry of 1,338 aa**, not the reviewed Q9P2D0 of 1,353 aa; E7EPI0 is simply the leading identifier of the MaxQuant protein group `E7EPI0;Q9P2D0;Q9P2D0-2;E9PDR5` [E1][E4]. And **IBTK isoform 2 is annotated Nucleus** while the canonical protein is cytoplasmic and membrane-peripheral [E4], so the group cannot tell you which isoform, or which compartment, the peptides came from. `DOT1L[tiab] AND IBTK[tiab]` and `(MLL OR KMT2A)[tiab] AND IBTK[tiab]` both return 0 [E5].

Go deeper
The reciprocal arm, which is the strongest evidence anywhere in this set

Hein et al. tagged 1,330 bait cell lines, and IBTK is one of them 3. That makes DOT1L-IBTK the only pair among the six with a within-study, within-cell-line, within-pipeline reciprocal test - a control that would normally take a separate experiment in a separate lab.

GFP-IBTK recovered 23 preys: POP7 and RPP25L (RNase P/MRP), EIF4A1, the RFC1/RFC3/RFC4 clamp loader, ribosomal proteins RPLP0, RPLP1, RPS3, RPS8, RPS9, RPS27 and RPL6, nucleolar factors DDX21, NAT10, TSR1 and WDR74, and MSL1, SRP72, LACTB, DHX30, ESPL1 and USP10 3. DOT1L is absent.

How much this is worth, stated honestly. AP-MS asymmetry is common and has mundane causes: the two baits carried the same LAP tag but at very different levels (IBTK 38,860 copies, DOT1L 3,734), and a prey at 1:476 sits at the detection floor whichever way round the experiment is done. The correct reading is not recovered in the reciprocal arm of this study. But the asymmetry cannot be used selectively - the same design that makes this negative weak would have made a positive strong, so a graph that counts the forward spoke as evidence has to count the reverse as evidence too.

The accession this project was about to key an entry on is unreviewed

The dispatch table, and analyses/directed-interactome-coverage-recheck's resolved-accession dictionary, both give IBTK as E7EPI0. That is what IntAct records, and IntAct records it because it is the first identifier in the MaxQuant protein group 1.

accessionreviewedlength
what IntAct recordsE7EPI0unreviewed (TrEMBL)1,338
the reviewed human proteinQ9P2D0reviewed (Swiss-Prot)1,353

Both are IBTK and neither is wrong, but an entry keyed on E7EPI0 would carry an unreviewed accession with no curated function, no curated subcellular location beyond automatic annotation, and a length 15 residues short of the canonical protein - and verify-accessions.py would pass it, because the accession and the length would be internally consistent 4. This entry asserts Q9P2D0 / 1,353 and records E7EPI0 as the source-database identifier in data.source_database_accession. The general form is written up in concept-a-protein-group-leading-id-is-not-the-interactor.

Why a B-lineage reader will want this to be real, and why that is a reason for care

IBTK inhibits BTK. BTK is the target of ibrutinib and the centre of B-cell receptor signalling, and this graph's index disease is infant B-lineage leukaemia. A DOT1L-IBTK link would be the most therapeutically suggestive of the six by a wide margin.

There is nothing behind it. Zero PubMed records pair IBTK with DOT1L or with MLL/KMT2A 5, the source evidence is the lowest-stoichiometry row in the pulldown, the reciprocal arm is negative, and the experiment was done in HeLa - a cervical carcinoma line with no B-cell receptor, no BTK signalling and no reason to assemble a BTK-regulatory complex at all. The one partner whose biology would matter here was measured in the one cell type where that biology does not exist.

Facts and figures
uniprotQ9P2D0
length AA1353
gene symbolIBTK
protein nameInhibitor of Bruton tyrosine kinase (IBtk)
uniprot resolutionaccession:Q9P2D0 AND organism_id:9606 AND reviewed:true; primary gene name read back as IBTK; length 1,353 aa checked. The unreviewed TrEMBL record E7EPI0 (1,338 aa) was fetched in the same query and deliberately NOT used as this entry's accession.
shown for fusionNO - nobody has looked in any fusion. Zero PubMed records pair MLL or KMT2A with IBTK.
source database accession
{
  "accession": "E7EPI0",
  "status": "unreviewed (TrEMBL)",
  "length_aa": 1338,
  "why": "leading identifier of the MaxQuant protein group E7EPI0;Q9P2D0;Q9P2D0-2;E9PDR5, carried through into IntAct and into analyses/directed-interactome-coverage-recheck",
  "hazard": "an entry keyed on it would be internally consistent and would pass verify-accessions.py, because accession and length agree with each other"
}
interaction with dot1l
{
  "source_pmid": "26496610",
  "source_class": "proteome-scale AP-MS atlas",
  "intact_interaction_ac": "EBI-11060394",
  "imex_id": "IM-24272-464",
  "detection_method": "psi-mi:MI:0007 anti tag coimmunoprecipitation",
  "interaction_type": "psi-mi:MI:0914 association",
  "expansion": "psi-mi:MI:1060 spoke expansion - binary pair inferred, not observed",
  "intact_miscore": 0.35,
  "author_confidence_class": "B",
  "core_interactor": false,
  "interaction_stoichiometry_prey_per_bait": 0.002099,
  "interaction_stoichiometry_as_ratio": "1 IBTK per 476 DOT1L - the lowest in the pulldown",
  "log10_prey_bait_ratio": -2.678,
  "abundance_stoichiometry_prey_per_bait": 9.832,
  "log10_prey_bait_expression_ratio": 0.9926466,
  "enrichment_sd": 2.287,
  "neg_log10_p": 4.022619,
  "prey_bait_correlation": 0.156,
  "valid_values": 3,
  "hela_copies_per_cell": 38317.12,
  "independent_sources": 0,
  "reciprocal_tested": true,
  "reciprocal_result": "NEGATIVE - GFP-IBTK (cell line MCP_ky_0004631, bait copy number 38,860) recovered 23 preys in the same study and DOT1L was not among them"
}
localization
{
  "subcellular": "Cytoplasm and Membrane (peripheral), translocating to the plasma membrane on IgM stimulation; ISOFORM 2 is separately annotated Nucleus, and Q9P2D0-2 is inside the protein group the DOT1L pulldown identified, so the compartment of the identified species is unresolved.",
  "source": "UniProt Q9P2D0, fetched 2026-09-21; the location comments carry no ECO codes in the returned fields (brief rule 91)"
}
absence claim
{
  "queries_run": [
    "DOT1L[tiab] AND (IBTK OR IBtk)[tiab]",
    "(MLL OR KMT2A)[tiab] AND (IBTK)[tiab]"
  ],
  "records": [
    0,
    0
  ],
  "operand_controls": {
    "DOT1L[tiab]": 710
  },
  "names_tried": [
    "IBTK",
    "IBtk (case-folded to the same term by PubMed)"
  ],
  "names_not_tried": [
    "KIAA1417",
    "BTKI",
    "\"inhibitor of Bruton tyrosine kinase\" as a phrase",
    "CUL3-IBTK as a complex name"
  ],
  "collision_registry_checked": "yes - data/symbol-collisions.json, 62 rows, grepped 2026-09-21; no row exists for IBTK, NIN, MAPK7, ERK5, PARP16, EBNA1BP2, EBP2 or BRINP1",
  "date": "2026-09-21",
  "tool": "PubMed E-utilities esearch",
  "bounded_by": "titles and abstracts only; supporting figures not searched"
}
How this could bite us back
  • A DOT1L-IBTK link points straight at BTK and therefore at ibrutinib, which is the single most tempting misreading available in this entry. The evidence is the lowest-stoichiometry row of one screen with a negative reciprocal arm, in a cell with no B-cell receptor.
  • Quoting E7EPI0 as this protein's accession propagates an unreviewed identifier into a graph whose accession checker cannot flag it, because the length matches.
  • If IBTK really does regulate BTK availability and also touches a chromatin methyltransferase, then IBTK depletion in a B-ALL model would confound a signalling readout with an epigenetic one - a reason to design any such experiment with an H3K79 control from the start.
Still open
  1. Does IBTK meet DOT1L in a cell that actually has a B-cell receptor? Everything here is HeLa, where BTK signalling does not exist, and a t(4;11) blast is the system where the question would mean something.
  2. IBTK isoform 2 is annotated nuclear and is inside the identified protein group. Which isoform was in the pulldown? An isoform-resolved reanalysis of the deposited peptides would answer it and would also resolve the compartment.
  3. How many other partner entries in this graph are keyed on an unreviewed TrEMBL accession inherited from a protein-group leading identifier? verify-accessions.py cannot detect this class because accession and length agree.
Every source behind this entry (5)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    The entire recorded interaction evidence for IBTK with DOT1L is one row of one IntAct interaction record - EBI-11060394 / IM-24272-464, spoke-expanded (MI:1060) from a single anti-tag co-immunoprecipitation of LAP-tagged DOT1L in HeLa, interaction type 'association' (MI:0914), intact-miscore 0.35, author-confidence B - from PMID 26496610, and the interactor accession IntAct records is E7EPI0, the leading identifier of the protein group E7EPI0;Q9P2D0;Q9P2D0-2;E9PDR5.

    What this evidence cannot show

    One database on one day. Spoke expansion means the binary pair was inferred by a curator from co-purification and never observed. The protein group is the unit the mass spectrometer resolved; picking any single accession out of it - including Q9P2D0 - is a choice, and the peptides may not distinguish the isoforms. HeLa only, no leukaemia cell, no fusion (brief rule 73).

    Evidence that argues against this (1)
    • Curating the leading group identifier is standard IMEx practice and is not an error - the alternative, curating four accessions as four interactors, would be worse. The defect is downstream, where a consumer reads one accession as 'the interactor' without noticing it is a group representative. EMBL-EBI IntAct 2026
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset

    In the source study's interactions table the DOT1L-IBTK spoke is class B with interaction stoichiometry 0.002099 (1 IBTK per 476 DOT1L, log10 prey:bait -2.678 - the lowest of the whole pulldown), abundance stoichiometry 9.832, enrichment 2.287 SD, -log10 p 4.023, prey-bait profile correlation 0.156 and core.interactor not set; IBTK is present in the paper's HeLa proteome at 38,317 copies per cell against DOT1L's 3,779.

    What this evidence cannot show

    Downloaded 2026-09-21 from the Elsevier static CDN with magic bytes checked (PK). Stoichiometries are medians of three replicates with no dispersion reported. The paper's class thresholds are defined in a methods section that is unreadable (no PMCID; the DOI resolves to a 13-character Elsevier stub), so the class is used as an ordinal grade published by the authors.

    Evidence that argues against this (1)
    • IBTK's prey-bait correlation of 0.156 is the highest of the six unwritten partners and is closer to AF9's 0.275 than to EBNA1BP2's 0.057, so on the one dimension that separates the A+ rows from the rest, IBTK is the least bad of the six rather than the worst. Cell 163(3), supplementary material 2015
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    IBTK was itself a bait cell line in the same study (MCP_ky_0004631, LAP tag, human, bait copy number 38,860); GFP-IBTK recovered 23 partners in the same HeLa background through the same pipeline - POP7, RPP25L, EIF4A1, RFC1, RFC3, RFC4, RPLP0, RPLP1, RPS3, RPS8, RPS9, RPS27, RPL6, DDX21, NAT10, TSR1, WDR74, MSL1, SRP72, LACTB, DHX30, ESPL1 and USP10 - and DOT1L is not among them.

    What this evidence cannot show

    A reciprocal AP-MS failure is weaker than it looks: tag position, the ten-fold difference in bait expression between the two cell lines, and a prey sitting at 1:476 all make asymmetric recovery common. This bounds the claim to 'not recovered in the reciprocal arm of this study' and not to 'does not interact'. The IBTK prey list was read from the same workbook and not independently verified against IntAct.

    Evidence that argues against this (1)
    • Both baits carry a C-terminal LAP tag, and a C-terminal tag on IBTK could occlude the same surface that a C-terminally tagged DOT1L needed - a shared-tag artefact would produce exactly this asymmetry for a reason that has nothing to do with whether the proteins bind. Cell 163(3), supplementary material 2015
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProtKB Q9P2D0 (IBTK_HUMAN) is the reviewed human IBTK, 1,353 aa, 'Inhibitor of Bruton tyrosine kinase', gene synonyms BTKI and KIAA1417, with subcellular location Cytoplasm and Membrane (peripheral), translocating to the plasma membrane on IgM stimulation, and a separate comment placing ISOFORM 2 in the Nucleus; E7EPI0, the accession IntAct and this project's coverage recheck record for the same row, is an UNREVIEWED TrEMBL entry of 1,338 aa with only automatic (ARBA) location annotation.

    What this evidence cannot show

    Both records fetched 2026-09-21 in one query with gene_primary and length requested and read back (brief rules 62 and 65). The 15-residue difference between them is far inside the range where length cannot discriminate (brief rule 69 part 2), so verify-accessions.py would pass either. The isoform-2 nuclear comment carries no ECO code in the returned fields and is therefore a curated statement, not a traced measurement (brief rule 91).

    Evidence that argues against this (1)
    • The nuclear isoform 2 cuts against the compartment-mismatch reading of this spoke: Q9P2D0-2 is IN the protein group the peptides mapped to, so a nuclear IBTK pool is annotated and present in the identification, and 'IBTK is cytoplasmic therefore it cannot meet DOT1L' does not survive reading the second location comment. UniProt Consortium 2026
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    Counted absence, PubMed E-utilities esearch run 2026-09-21: DOT1L[tiab] AND (IBTK OR IBtk)[tiab] returns 0 records and (MLL OR KMT2A)[tiab] AND IBTK[tiab] returns 0 records, with the operand control DOT1L[tiab] returning 710.

    What this evidence cannot show

    Bounded by titles and abstracts (brief rule 80). PubMed case-folds, so IBTK and IBtk are the same query and the querytranslation showed both limbs collapsing to ibtk[All Fields]; the retired-style alias KIAA1417 and the spelled-out name were not run. IBTK is a short symbol of the shape brief rule 88 warns about, but no row for it exists in data/symbol-collisions.json and a bare search returns BTK-inhibitor biology rather than another field's vocabulary.

    Evidence that argues against this (1)
    • Searching only IBTK misses any paper framing the work at the level of the CUL3-IBTK ubiquitin ligase or of BTK regulation without naming the gene - the complex-name problem brief rule 64 documents - so this zero bounds the symbol, not the biology. NCBI PubMed (queries run 2026-09-21) 2026
Connected to