ASH1L

Summary: ASH1L binds DNA through chromatin-associated reader modules.

Reason: The electronic AT-hook-based inference agrees with direct human ASH1L DNA-binding measurements reported for its bromodomain and BAH region in PMID:40044670. These are biochemical DNA interactions, not evidence that ASH1L is a sequence-specific transcription factor.

Summary: Chromatin binding recruits and positions the ASH1L histone writer.

Reason: Human ASH1L occupies active chromatin in PMID:17923682. PMID:40044670 independently dissects H3K4me2/3 recognition by its PHD and DNA recognition by adjacent modules. Recognition of methylated H3K4 is not catalytic methylation of that residue.

Summary: Nuclear localization is independently established for human ASH1L.

Reason: Accept the compartment on independent human immunostaining and chromatin evidence, while deferring the target-specific assay in the original 2015 screen to its curator. The cached abstract of PMID:25593309 foregrounds ZMYND8; that does not establish that ASH1L was untested or misattributed.

Summary: ASH1L is a nuclear chromatin enzyme.

Reason: The electronic localization agrees with human immunostaining and chromatin occupancy. Domain identity alone is not being used to infer every nuclear subcompartment.

Summary: The original human study directly reports intranuclear ASH1L.

Reason: The original paper abstract describes immunostaining with several anti-huASH1 antibodies and intranuclear speckles. This supports the seeded TAS nuclear localization without converting its evidence code to IDA.

Summary: The inherited nucleoplasmic localization agrees with target evidence.

Reason: The PAINT localization agrees with independent human nuclear immunostaining and the supported nucleoplasmic Human Protein Atlas annotation.

Summary: Human Protein Atlas supports nucleoplasmic ASH1L.

Reason: The ASH1L-specific Human Protein Atlas record reports nucleoplasm with supported reliability. Its uncertain Golgi signal does not establish an additional compartment.

Summary: Nucleoplasm is consistent with ASH1L chromatin catalysis.

Reason: Independent human localization supports this existing compartment while the Reactome-specific attribution is deferred to its curator. The cached R-HSA-4827383 summary names other methyltransferases and does not itself identify ASH1L; this summary limitation is not proof of a wrong-gene annotation.

Summary: Nucleoplasm is the supported compartment for ASH1L histone methylation.

Reason: Human target localization corroborates this compartment. R-HSA-5638157 describes ASH1L-mediated H3K36 dimethylation, although the short cached reaction summary does not independently spell out the compartment.

Summary: ASH1L associates with chromosomal chromatin.

Reason: Human chromatin immunoprecipitation in PMID:17923682 supports association with transcribed chromosomal regions. The general chromosome label is compatible with its chromatin-writing function and does not imply mitotic kinetochore function.

Summary: Tight-junction localization is reported but its functional role is unresolved.

Reason: The original human immunostaining study supports localization at intercellular tight junctions. A junctional catalytic or adhesion-signaling function has not been established here; retain the location outside the nuclear catalytic core.

Summary: The original human study supports a secondary tight-junction pool.

Reason: The abstract explicitly describes double immunofluorescence with junctional markers. Its suggestion of adhesion-mediated signaling is a hypothesis, not a demonstrated biological process to add.

Summary: Histone methylation is a direct contribution to chromatin remodeling in the current GO scope.

Reason: ASH1L itself methylates nucleosomal histone H3. GO:0006338 includes histone-modifying activity through the ontology part_of relationship; this annotation does not assert ATP-dependent nucleosome sliding. The valid K36 activity supports the process despite unresolved K4/K9 mappings among the electronic inputs.

Summary: ASH1L regulates transcription through chromatin modification.

Reason: Human chromatin occupancy and target-gene perturbation data support regulation of transcription through histone modification, consistent with this ancestral-function assertion.

Summary: HPA supports localization in nuclear bodies without defining a separate core activity.

Reason: The current ASH1L target record lists nuclear bodies as an additional supported localization. The original human study also describes intranuclear speckles; neither establishes a specific body-dependent catalytic mechanism.

Summary: Nuclear lumen is compatible with the established nuclear pool.

Reason: This broad electronic compartment agrees with direct human intranuclear staining and nucleoplasmic localization. It is retained as a valid source annotation; the core is localized more precisely to nucleoplasm.

Summary: Specify the supported H3K36 dimethyltransferase activity.

Reason: The generic histone methyltransferase inference is correct but under-specific. Human enzyme assays in PMID:26002201 distinguish mono- and dimethylation of nucleosomal H3K36 from SETD2-dependent trimethylation. Use the supported specific product activity.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: Conflicting primary substrate assignments prevent confident adjudication of this IBA.

Reason: Historical human and mouse studies report ASH1L H3K4 activity (PMID:17923682; PMID:24012418), whereas human substrate-mutant assays in PMID:26002201 support K36 specificity and no K4 activity. The conflicting assays and the actual PAINT node placement remain unresolved. Human PHD recognition of methylated H3K4 in PMID:40044670 does not establish K4 catalysis or settle the ancestral-function assertion.

Summary: Automatic transfer of H3K4 writer activity is excessive in light of direct human substrate-specificity assays.

Reason: Human ASH1L assays in PMID:26002201 show H3K36 mono- and dimethylation, loss of activity on H3K36A nucleosomes, and no activity with H3K4 as substrate. PMID:40044670 separately establishes recognition of pre-existing H3K4me2/3 by the PHD and H3K36 methylation by SET-containing constructs. The mouse-to-human transfer therefore overstates H3K4 catalysis. Historical H3K4 claims in PMID:17923682 and PMID:24012418 remain disputed, not disproven assay errors; construct and substrate differences require reconciliation.

Summary: H3K9 catalytic activity cannot yet be resolved for human ASH1L.

Reason: The mouse Ash1l-specific assay underlying PMID:22939622 remains unverified. Its Prdm3/Prdm16-focused abstract cannot establish whether Ash1l was tested. Human K36 specificity does not exclude all alternative substrate or cofactor contexts, so this H3K9 assertion remains unresolved.

Summary: The curated ortholog transfer requires the unread donor assay.

Reason: The mouse Ash1l-specific assay underlying PMID:22939622 remains unverified. Its Prdm3/Prdm16-focused abstract cannot establish whether Ash1l was tested. Human K36 specificity does not exclude all alternative substrate or cofactor contexts, so this H3K9 assertion remains unresolved.

Summary: Represent the experimentally supported H3K36 mono- and dimethylation reaction.

Reason: PMID:26002201 reports ASH1L mono- and dimethylation, H3K36A loss of activity, and SETD2-only trimethylation in vitro. GO:0140954 defines successive addition of two methyl groups starting from unmethylated H3K36 and includes histone H3K36 mono/dimethylase activity as a synonym. It therefore represents both successive methyl transfers supported by these assays.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: Refine the K36 inference to the supported mono/dimethylation reaction.

Reason: The K36-specific inference agrees with direct human assays in PMID:26002201. GO:0140954 covers successive mono- and dimethylation of initially unmethylated H3K36, matching the measured product states.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: Replace broad H3 methyltransferase activity with K36 dimethylation.

Reason: Independent human assays establish the narrower activity. The short original Reactome cache does not name ASH1L, so it is not presented as independent target-level experimental proof; its original TAS source is preserved while the biological refinement relies on PMID:26002201.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: The reaction supports the specific H3K36 dimethyltransferase activity.

Reason: R-HSA-5638157 names ASH1L and methylation of the monomethylated lysine to dimethyllysine, consistent with human kinetic assays. Its residue37 notation counts the initiator methionine; the conventional histone site is H3K36.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: The H3K9 monomethylation mapping needs independent human support.

Reason: The mouse Ash1l-specific assay underlying PMID:22939622 remains unverified. Its Prdm3/Prdm16-focused abstract cannot establish whether Ash1l was tested. Human K36 specificity does not exclude all alternative substrate or cofactor contexts, so this H3K9 assertion remains unresolved. The EC/Rhea mapping supplies no separate experiment.

Summary: The specific H3K9 monomethylation transfer remains unresolved.

Reason: The mouse Ash1l-specific assay underlying PMID:22939622 remains unverified. Its Prdm3/Prdm16-focused abstract cannot establish whether Ash1l was tested. Human K36 specificity does not exclude all alternative substrate or cofactor contexts, so this H3K9 assertion remains unresolved.

Summary: Retain K36 methyltransferase activity with supported dimethyl product specificity.

Reason: PMID:21239497 assigns the additional trimethylated product to the hyperactive Q2265A mutant. Independent human enzyme assays in PMID:26002201 show ASH1L mono- and dimethylation, with trimethylation only for SETD2 under the tested conditions. GO:0140954 represents the successive two methyl transfers supported for ordinary ASH1L; the mutant-specific trimethylation result remains valid in its experimental context.

Proposed replacements: histone H3K36 dimethyltransferase activity

Summary: The trimethyltransferase mapping overextends mutant activity to the ordinary protein.

Reason: PMID:21239497 assigns the additional trimethylated product to the hyperactive Q2265A mutant. Independent human enzyme assays in PMID:26002201 show ASH1L mono- and dimethylation, with trimethylation only for SETD2 under the tested conditions. GO:0140954 represents the successive two methyl transfers supported for ordinary ASH1L; the mutant-specific trimethylation result remains valid in its experimental context.

Proposed replacements: histone H3K36 dimethyltransferase activity