BAHCC1 - a BAH-domain H3K27me3 reader that an MLL fusion binds and upregulates in mouse, and the locus carrying the two largest MLL1-N-reactive rises under a menin inhibitor in human RS4;11

No three-dimensional model is available for this protein, so the map above is the whole picture. That is itself informative: AlphaFold does not model chains this long, and the parts that have been crystallised are held under the entries for the individual domains.

• BAHCC1's tandem Tudor domain loads MCM at replication origins, so a BAHCC1-directed agent is a replication-licensing agent as well as a polycomb-reader agent. The essential-function toxicity of that is not a leukaemia question - every proliferating tissue in a child licenses origins - and it would not be visible in a screen scored on leukaemic cell death.
• The published leukaemia rationale is partner-dependent inside the KMT2A family itself (BAHCC1 lower in MLL-AF9 than in other MLL-fusion AML). A trial enrolling 'KMT2A-rearranged' as a single stratum would mix a plausible responder group with a plausible non-responder group and could return a null that is really a stratification failure - the same shape as the partner-outcome entries this graph already holds.
• In neuronal cells BAHCC1 is a positive regulator that antagonises SIN3A-HDAC1, and its loss impairs neurite outgrowth. A systemic BAHCC1 inhibitor in an infant therefore has a named CNS liability with a published mechanism behind it, not a hypothetical one.

1. Is the 4.036 rise at chr17:79,370,000 recruitment or renormalisation? The MENIN channel at the same bin is 0.056, and a within-sample share rises by construction where signal is lost elsewhere. Scoring the same two bins against the RS4;11 input and in the MOLM13 comparator - both already on disk from Q118 - settles it in an afternoon and has not been done.
2. What is at chr17:79,314,000? It carries the second-largest MLL1-N rise in the top thirty and overlaps no annotated gene on the GRCh37 primary assembly. If it is a distal element contacting the BAHCC1 promoter, the two bins are one observation; if not, they are two. Micro Capture-C in KMT2A::AFF1 patient material already exists for other loci (PMID 40729681) and could be asked this directly.
3. Does BAHCC1 protein exist, and is it H3K27me3-engaged, in KMT2A::AFF1 pro-B blasts? Nothing in the literature measures BAHCC1 protein in B-lineage KMT2A-rearranged leukaemia, and the one MLL-fusion result is mouse myeloid MLL-ENL with an explicit partner-dependence.
4. Which reader module is doing the work at any BAHCC1-occupied site - the H3K27me3-reading BAH domain or the H4K20me1-reading tandem Tudor domain? The two make opposite predictions about what a BAHCC1 inhibitor would do to a proliferating blast, and no published experiment separates them at a locus.
5. Does this graph need a mark-h4k20me1 entry? It holds mark-h4k20me2 and nothing for the mono-methyl state, so a documented reader activity of a gene in this beat has nowhere to link.

1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   BAHCC1 is Q9P281 (BAHC1_HUMAN), 2,639 aa, with gene synonyms BAHD2 and KIAA1447; its feature table annotates a single BAH domain at residues 2513-2633 plus fifteen disordered regions, and it carries no FUNCTION comment and no SUBCELLULAR LOCATION comment.

   UniProtKB Q9P281 (BAHC1_HUMAN), BAH and coiled-coil domain-containing protein 1 · UniProt Consortium · UniProt 2026
   ▸What this evidence cannot show

   Fetched live 2026-09-23 via gene_exact:BAHCC1 AND organism_id:9606 AND reviewed:true, which returned exactly ONE reviewed record, so the rule-62 several-reviewed-records trap does not fire. data/symbol-collisions.json (208 rows) was word-boundary grepped for BAHCC1 before the first query and has no row. The absence of FUNCTION and SUBCELLULAR LOCATION comments is a statement about the database on this date, not about the protein: both published reader activities are missing from it, and so is the tandem Tudor domain.

   ▸Evidence that argues against this (1)

   • The feature table is incomplete in a way that changes the biology: the H4K20me1-reading tandem Tudor domain characterised in 2025 is not annotated, so a domain-level analysis built from UniProt alone sees only the polycomb half of this protein. Rule 91's point generalises - a database's own annotation is a pointer, not a source. Nat Commun 2025
2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   HGNC assigns BAHCC1 (HGNC:29279, 'BAH domain and coiled-coil containing 1') the primary Ensembl gene id ENSG00000266074, cytoband 17q25.3 and UniProt Q9P281, with alias symbols KIAA1447 and BAHD2 and no previous symbol; Ensembl places ENSG00000266074 on GRCh38 at chr17:81,395,450-81,466,331 on the plus strand.

   HGNC record HGNC:29279 (BAHCC1) and Ensembl gene ENSG00000266074 · HUGO Gene Nomenclature Committee; Ensembl · genenames.org / Ensembl 2026
   ▸What this evidence cannot show

   Both records fetched live 2026-09-23 and resolved before any sentence using them was written. HGNC lists no previous symbol for BAHCC1, so the rule-69 retired-symbol undercount does not apply and no second literature arm was needed. The span given is the GRCh38 primary assembly; see E7 for why the GRCh37 span cannot be retrieved the same way.

   ▸Evidence that argues against this (1)

   • An HGNC record supplies ONE primary Ensembl id, and that is the point of using it (rule 78) - but it is also its limitation: about one gene in eleven in this corpus multi-maps to several Ensembl ids, and for BAHCC1 the primary id does not exist on the GRCh37 primary assembly at all. A join built on this row alone would silently fail against any hg19 table, which is exactly what E7 found. Ensembl GRCh37 archive 2026
3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   A conserved BAH module of BAHCC1 reads H3K27me3 specifically through a hydrophobic trimethyl-lysine cage, colocalising BAHCC1 with H3K27me3-marked genes and enforcing their silencing; BAHCC1 is highly expressed in human acute leukaemia and interacts with transcriptional corepressors, and depleting BAHCC1 or disrupting the BAH-H3K27me3 interaction derepresses H3K27me3-targeted tumour-suppression and differentiation genes and suppresses oncogenesis.

   BAHCC1 binds H3K27me3 via a conserved BAH module to mediate gene silencing and oncogenesis · Fan H, Lu J, Guo Y, Li D, Zhang ZM, Tsai YH, et al. · Nat Genet 2020
   PMID 33139953DOI
   ▸What this evidence cannot show

   Abstract-level read; the full text was not opened this session. 'Human acute leukemia' in the abstract is not resolved to a lineage or a fusion, and no KMT2A-rearranged or B-lineage experiment is named at abstract level. A germline Bahcc1 mutation disrupting H3K27me3 engagement causes partial postnatal lethality in mice, so this is a developmental function and not a leukaemia-restricted one.

   ▸Evidence that argues against this (2)

   • The pure-silencing model does not hold in every lineage. In neuronal cells BAHCC1 sits mainly at H3K27ac-marked regulatory elements, co-occupies them with SIN3A but not SIN3B, and ANTAGONISES SIN3A-HDAC1 deacetylation, so reducing it lowers acetylation and expression of proximal genes. BAHCC1 is a positive regulator there - the opposite sign from the leukaemia result. Monziani A et al. 2025
   • BAHCC1 expression can be driven by a super-enhancer independently of any polycomb input, in melanoma, where it supports proliferation and genome stability. High BAHCC1 in a tumour is therefore not evidence that an H3K27me3-reading mechanism is operating in it. Cell Rep 2023
4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   MLL-ENL upregulates Bahcc1 by binding its promoter in mouse haematopoietic stem and progenitor cells; Bahcc1 contributes to MLL-ENL-mediated immortalisation at least partly through repression of the H3K27me3-marked Cdkn1c; depletion of Bahcc1 suppresses MLL-ENL leukaemogenic activity in bone-marrow transplantation; and in public data high BAHCC1 expression is associated with poor prognosis in paediatric AML, with BAHCC1 significantly LOWER in MLL-AF9 AML than in other MLL-fusion AML.

   Bahcc1 is critical for the aberrant epigenetic program in a mouse model of MLL-ENL-mediated leukemia · Nakamura A, Masuya M, Shinmei M, Tawara I, Nosaka T, Ono R · Blood Adv 2024
   PMID 38452334DOI
   ▸What this evidence cannot show

   Abstract-level read; the full text was not opened this session. MOUSE, MYELOID and MLL-ENL (KMT2A::MLLT1) throughout - not human, not B-lineage, not KMT2A::AFF1, and the system is named in the claim itself rather than only here (rule 73). The prognostic association is a public-database correlation, not an experiment. The paper's own observation that BAHCC1 is lower in MLL-AF9 than in other MLL-fusion AML establishes that this axis varies BY PARTNER inside the KMT2A family.

   ▸Evidence that argues against this (1)

   • The same abstract supplies the counter-evidence to its own generalisation: within MLL-fusion AML, BAHCC1 is significantly lower in MLL-AF9 than in the other fusions. A partner-dependent axis is one whose absence in KMT2A::AFF1 would be unremarkable, so a negative result in the index fusion would not refute the mouse finding - and that asymmetry has to be stated before anyone runs the experiment, not after. Blood Adv 2024
5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   An evolutionarily conserved tandem Tudor domain of BAHCC1 selectively reads H4K20me1 and promotes recruitment of BAHCC1 and the MCM complex to replication origins; depleting BAHCC1 or disrupting the TTD:H4K20me1 contact reduces H4K20me1 levels and MCM loading and causes defects in replication-origin activation and cell-cycle progression.

   BAHCC1 binds H4K20me1 to facilitate the MCM complex loading and DNA replication · Li D, Zhang ZM, Mei L, Yu Y, Guo Y, Mackintosh SG, et al. · Nat Commun 2025
   PMID 40592879DOI
   ▸What this evidence cannot show

   Abstract-level read; the full text was not opened, so the residue boundaries of the tandem Tudor domain were NOT obtained. They are deliberately absent from data.domains rather than guessed, because a domain lane with an invented coordinate draws a false figure and is worse than a missing one. No cell system relevant to this project is named at abstract level. This graph holds mark-h4k20me2 and has no H4K20me1 entry, so this reader activity currently has nowhere in the graph to link.

   ▸Evidence that argues against this (1)

   • A replication-linked reader function means BAHCC1 chromatin occupancy is not a polycomb readout by default. A ChIP signal at a BAHCC1-bound site - including the one this beat is about - could reflect H4K20me1/origin biology rather than H3K27me3 silencing, and the two are indistinguishable without the mark measured alongside. Nat Genet 2020
6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

   In RS4;11 (KMT2A::AFF1), the hg19 2 kb bin chr17:79,370,000 reaches 128.3x the MLL1-N track's own genome median and is in the top thirty bins genome-wide; under VTP-50469 its MLL1-N within-sample share ratio is 4.036 with MENIN at 0.056, and the bin at chr17:79,314,000 carries 3.407 with MENIN at 0.078 - the two largest per-bin MLL1-N rises in that top thirty.

   Q118 re-analysis of GSE127507, instrument B: 2 kb genome-wide bins from the deposited .deduped.tdf files (RS4;11 MLL1n GSM3635695/96, MENIN GSM3635693/94, input GSM3635699); analyses/q118-menin-at-the-human-index-fusion/results-instrumentB.json, run 2026-09-23 · Iqbal Lab oncofusion knowledge graph 2026
   ▸What this evidence cannot show

   THE MEASUREMENT THAT NAMED THIS GENE IS NOT FUSION-SPECIFIC. The track is MLL1n, Bethyl A300-086A, an MLL1 N-TERMINUS antibody; the N-terminus is retained in KMT2A::AFF1, so the track reads the fusion and wild-type MLL1 together in unknown proportion, at a masking factor the same beat measured as 3.54-4.00x. Every occupancy statement drawn from it is about MLL1-N-reactive material and never about the fusion. n = 1 per antibody per condition - GSE127507 has no replicate anywhere - and the depositors measured a Drosophila spike-in and deliberately did not apply it, so every cross-condition figure is a within-sample share or rank, not a fold change. A per-bin share ratio from an n=1 comparison at a single 2 kb bin is the noisiest quantity in that beat, and a within-sample share RISES by construction wherever total signal is being lost elsewhere in the same sample, so 4.036 is not by itself evidence of new recruitment.

   ▸Evidence that argues against this (1)

   • In the channel that actually is evicted - MENIN - nothing at this locus goes up: the share ratios are 0.056 and 0.078, among the lowest in the top thirty. The rise is confined to an antibody the same beat measured as barely evicted globally (held share ratio 0.9069, per-bin median 0.8196), which is the signature of a renormalisation artefact as much as of relocation. Iqbal Lab oncofusion knowledge graph 2026
7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

   On the GRCh37 PRIMARY assembly the 2 kb bin chr17:79,370,000 falls inside the BAHCC1 5' region - ENSG00000263053 spans 79,369,275-79,375,658 (+) and ENSG00000171282, the unnamed GRCh37 model of BAHCC1, spans 79,373,540-79,433,357 (+) - while the bin at chr17:79,314,000 overlaps NO annotated feature of any biotype, the nearest being TMEM105 ending at 79,304,474 and a lincRNA beginning at 79,336,072. ENSG00000266074 itself is not on the GRCh37 primary assembly; the GRCh37 REST lookup returns it on the patch scaffold HG271_PATCH.

   Ensembl GRCh37 REST: overlap/region/human/17:79300000-79450000?feature=gene, and lookup/id/ENSG00000266074 · Ensembl · Ensembl GRCh37 archive 2026
   ▸What this evidence cannot show

   Fetched live 2026-09-23. This is a COORDINATE query, not a name lookup, so it cannot make the UCSC longest-match error that cost Q117 time. It resolves GENE MODELS only: 'no annotated feature' means no gene, not no function, and an unannotated 2 kb interval 56 kb upstream of a gene is the expected shape of a distal enhancer that this query cannot see. The deposited data is hg19 and Ensembl GRCh37 gene models are not the RefGene models the depositors' own ROSE windows were built from, so the two annotations are not guaranteed to agree bin-for-bin.

   ▸Evidence that argues against this (1)

   • This check disagrees with the PROSE of the entry that named the gene, which reports both bins as being 'at BAHCC1'. It agrees with that same entry's open_questions field, which lists chr17:79,314,000 among four top-thirty bins in no RefGene TSS window. The disagreement is internal to that entry and is reported to the coordinator rather than edited here. Iqbal Lab oncofusion knowledge graph 2026
8. E8moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   In neuronal-lineage cells BAHCC1 associates mainly with H3K27ac-marked regulatory elements, co-occupies them with the SIN3A scaffold but not its paralog SIN3B, and antagonises SIN3A-HDAC1 deacetylation; reducing Bahcc1 lowers acetylation and expression of proximal genes enriched for neurogenesis and cell migration, and impairs neurite outgrowth in primary cortical neurons.

   BAHCC1 promotes gene expression in neuronal cells by antagonizing SIN3A-HDAC1 · Monziani A, Perry RB, Hezroni H, Ulitsky I · Nucleic Acids Res 2025
   PMID 40682820DOI
   ▸What this evidence cannot show

   Abstract-level read. Neuronal lineage, not haematopoietic. Cited as the published counter-model to a repression-only reading of BAHCC1, not as evidence about leukaemia.

   ▸Evidence that argues against this (1)

   • The leukaemia result is the counter-evidence here, and for this project's purposes it is the better-matched one: in leukaemia, depleting BAHCC1 DEREPRESSES H3K27me3-targeted genes, which is repression, not activation. Both can be true in different lineages and this entry does not resolve them. Nat Genet 2020
9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

   BAHCC1 is nearly absent from the KMT2A and lymphoblastic literature: BAHCC1[tiab] returns 23 records, BAHCC1[tiab] AND KMT2A[tiab] returns 2 (both unrelated to fusion leukaemia - an intestinal-stem-cell paper and a neuronal Kmt2a deletion paper), BAHCC1[tiab] AND MLL[tiab] returns 1, BAHCC1[tiab] AND leukemia[tiab] returns 3, BAHCC1[tiab] AND H3K27me3[tiab] returns 4, and both BAHCC1[tiab] AND "acute lymphoblastic"[tiab] and BAHCC1[tiab] AND "MLL-AF4"[tiab] return 0.

   PubMed esearch via scripts/pubmed_safe.py, 2026-09-23 · Q119-genes-the-measurement-named · NCBI PubMed 2026
   ▸What this evidence cannot show

   Bounded by TITLES AND ABSTRACTS only (rule 80): a BAHCC1 measurement inside a supporting figure of a KMT2A paper is invisible to every one of these queries. Each term carries its own [tiab] tag and no parenthesised group was used (rule 94); every querytranslation was read back from the response and each expanded to "TERM"[Title/Abstract] as written. The two KMT2A hits and the three leukemia hits WERE read at title level and are characterised above; nothing else in these sets was read, so these are counts and not a review.

   ▸Evidence that argues against this (1)

   • A near-zero literature count here is not evidence of biological irrelevance; it is evidence of a young gene. BAHCC1's first reader-function paper is 2020 and its second is 2025, so the functional corpus is five years old and the KMT2A intersection has had one substantive paper in it since 2024. Blood Adv 2024

• binds →H3K27me3 - the repressive mark this graph argued about for seventy-seven cycles without ever writing an entry for it

  BAH module (2513-2633) reads H3K27me3 via a trimethyl-lysine cage

• regulates →KMT2A::MLLT1 (MLL-ENL), t(11;19)(q23;p13.3)

  MLL-ENL binds the Bahcc1 promoter and upregulates it; Bahcc1 depletion suppresses MLL-ENL leukaemogenesis (MOUSE, myeloid)

• related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)

  locus carrying the two largest per-bin MLL1-N-reactive rises under VTP-50469 in RS4;11 - NOT a fusion-specific measurement

• supports →Instrument check: the top thirty RS4;11 MLL1-N 2 kb bins recovered from a deposited .tdf are JMJD1C (259x the genome median), ZMYND11, BAHCC1, CDK6, MYB, RUNX2, IGF2BP2, FOXP1, SATB1 and ARID1B - canonical KMT2A::AFF1 targets, annotated only against the series' own RefGene windows

  the entry that named this gene; this entry corrects its annotation of the second bin

• depends on →PRC2 - the complex, not the subunit, is the unit every experiment in this disease actually resolves

  reads the PRC2 product H3K27me3; BAHCC1 occupancy is downstream of PRC2 deposition

• related to →Revumenib (SNDX-5613, Revuforj) — first approved menin–KMT2A interface inhibitor

  its promoter bin is where MLL1-N-reactive share rises most under the menin inhibitor VTP-50469, the tool compound of this drug class

• binds →H4K20me1 - the transcription- and replication-linked H4 state, whose gene-body level tracks Pol II; it HAS been profiled genome-wide in human haematopoietic stem cells twice, and a query scoped to 'leukemia' cannot see either dataset

  tandem Tudor domain reads H4K20me1 SELECTIVELY over me2, recruiting BAHCC1 and the MCM complex to replication origins; BAHCC1 loss also LOWERS H4K20me1 - this is the link that had nowhere to point before this node existed