The normal seed contains 33 GOA assertions and the two recorded products Q8IXJ9-1 and Q8IXJ9-2. Their identifiers, qualifiers, evidence codes, references and WITH/FROM entities are preserved. The current provisional synthesis has 12 ACCEPT, 10 KEEP_AS_NON_CORE, 5 MODIFY and 6 UNDECIDED decisions, with no NEW or REMOVE assertions. Final authorship awaits inspection of the five additional normal publication caches. The seeded YAML remains unchanged at this stage.
ASXL1 is a noncatalytic regulator of PR-DUB. PMID:20436459(https://pmc.ncbi.nlm.nih.gov/articles/PMC3182123/) reconstitutes human BAP1 with human ASXL1 residues 2–365 and demonstrates ASXL1-dependent removal of nucleosomal H2A ubiquitin. The human experiments are distinct from the paper's Drosophila genetics and Calypso catalytic mutant. Targeted human-complex Results and purification/activity Methods were read. This supports refining the corresponding generic BAP1-binding annotation to deubiquitinase activator activity, GO:0035800.
The final structural study PMID:37556531(https://pmc.ncbi.nlm.nih.gov/articles/PMC10411902/) examines human BAP1 and the ASXL1 DEUBAD fragment, with structural and turnover substrates distinguished. ASXL1 contacts nucleosomal DNA and positions the complex for BAP1-catalyzed hydrolysis. Selected Results, construct/activity/binding Methods and captions were independently assessed; supplementary data and figure pixels were not inspected. DNA contact mutations affect catalysis without necessarily changing equilibrium affinity. This is evidence of a regulatory contribution, not an autonomous ASXL1 ubiquitin-hydrolase active site. The earlier preprint is not substituted for the final paper.
The second proposed core is transcriptional coactivation, GO:0003713. PMID:16606617(https://pubmed.ncbi.nlm.nih.gov/16606617/) supports receptor association and cooperation with SRC-1. An independent read of original Figure 5A identifies human ASXL1 residues 1096–1541 tested against human ESR1, mouse RXRalpha and chicken THRB. Figure 5B uses full-length mouse ASXL1. The three GOA partner accessions remain unchanged. Increased promoter histone acetylation does not establish ASXL1 acetyltransferase activity. Broad human chromatin/transcriptional evidence in PMID:30664650(https://pmc.ncbi.nlm.nih.gov/articles/PMC6341105/) also informs this core; later double-knockout experiments and full Methods require further reading.
PMID:22897849(https://pmc.ncbi.nlm.nih.gov/articles/PMC3422511/) contains human BAP1, EED and EZH2 co-association evidence. Targeted Results, immunoprecipitation Methods and Figure 2/6 captions were read. Cellular co-immunoprecipitation supports the reported associations without proving purified binary affinity. The differing HOXA effects of BAP1 and ASXL1 depletion also caution against treating PRC2-associated repression as identical to PR-DUB catalysis. These three measured generic associations are provisionally NON_CORE under the user-supplied action definitions; the standing generic-binding policy clarification remains unanswered.
PMID:21047783(https://pmc.ncbi.nlm.nih.gov/articles/PMC3020743/) supports contextual PPAR regulation and adipocyte phenotypes. Its expression constructs are mouse ASXL1 and human ASXL2. Figure 1C instead detects endogenous ASXL1 in human HEK293 cells with transfected FLAG-PPARgamma. The species and expression boundaries are retained. Methods and selected Results/captions were read; figure pixels and supplements were not.
Six interaction assertions remain unresolved at the exact source/construct level: AKT1 in PMID:18624398 and PMID:35512704; BAP1 in PMID:19615732, PMID:27705803 and PMID:33961781; and GOLGA6L9 in PMID:32296183. The fetched UniProt record positively corroborates the partners. Missing access to a target-specific table does not show that an interaction is false. The 33961781 and 35512704 caches contain substantive partial HTML, not complete experimental accounts. Their actual cache availability flags will be retained separately from these reading limits.
The official GO change #32520, merged 2026-08-24, obsoletes GO:0042974 and several receptor-specific binding terms. Its suggested current term is GO:0016922, nuclear receptor binding. The review preserves the obsolete source ID and proposes the current replacement; it does not treat a consider mapping as automatic equivalence. GO:0042975 was not included in this change.
PAINT node placement was not independently reconstructed. The existing IBA assertions are evaluated against target biology while source-placement uncertainty is recorded separately. A short donor list, an ASXL2 donor, or ASXL1 appearing in its own WITH/FROM is not evidence of a transfer error.
The HP1/LSD1 paper PMID:19880879 is retracted, as confirmed by PMID:25750265(https://pubmed.ncbi.nlm.nih.gov/25750265/) and UniProt's caution. Its reported interactions are not used as established ASXL1 functions. This does not automatically invalidate other papers by the same authors. No new developmental, cancer, methylated-DNA recognition or histone-catalytic annotation is proposed from these preliminary readings.
Bohring-Opitz syndrome is documented in the fetched UniProt record. The complete official abstract of PMID:21706002(https://pubmed.ncbi.nlm.nih.gov/21706002/) independently reports de novo ASXL1 nonsense variants in affected individuals; the full clinical paper was not read. This supplies disease context without assigning an additional molecular function.
The normal Falcon research attempt with its configured fallback returned nonzero and produced no provider report. Manual research is recorded here. The subsequent PMID cache check succeeded and preserved all existing files. Four previously absent normal caches were imported with the seed; six pre-existing caches were retained byte for byte. PMID:19615732 was unavailable to the remote seed fetcher, but an existing canonical abstract cache supplies its actual title. Only the authored reference title will be corrected; downloaded source files remain immutable.
The indexed primary text of PMID:30664650(https://pmc.ncbi.nlm.nih.gov/articles/PMC6341105/) subsequently exposed Results and legends for Figures 1, 2 and 5, plus the Discussion. Human HAP1 single knockouts have modest effects; combined ASXL1/ASXL2 loss reduces BAP1 protein and increases H2A ubiquitination, with predominantly reduced transcript levels. ASXL3 is not expressed in this model. This supports context-specific redundancy, not universal interchangeability. HeLa affinity purification distinguishes the stable BAP1–ASXL core from associated factors; KDM1B is preferentially associated with ASXL2. The transcriptional coactivator interpretation also uses human ASXL1/PolII chromatin profiles, separately from mouse BAP1/PolII data. These findings strengthen the proposed coactivation role without making every transcriptional effect independent of PR-DUB. Full Methods, supplementary datasets and figure pixels remain unread. Direct publisher and PMC opens were challenged; the readable evidence was the indexed original text.
The five additional publications are now exact normal-cache imports. PMID:21047783 remains abstract-only; the four other records contain XML-derived full text. Selected recovered Results and Methods were inspected as recorded below. This supersedes the acquisition-pending statements above. All 33 original annotation objects and both products remain unchanged. Final decisions are 12 ACCEPT, 10 KEEP_AS_NON_CORE, 5 MODIFY and 6 UNDECIDED; no new assertion is added. The two molecular cores are BAP1 activation and transcriptional coactivation. Source-specific uncertainty and the existing generic-binding policy question remain explicit.
All 33 original source assertions, 33 actions and both alternative products are retained: 12 ACCEPT, 10 KEEP_AS_NON_CORE, 5 MODIFY and 6 UNDECIDED. No NEW annotation is introduced. The second core is narrowed to ligand-dependent RAR/RXR coactivation with SRC-1; transcriptional effects of BAP1/ASXL1/2 loss remain integrated with the PR-DUB context. The 2006 receptor experiments do not establish BAP1 or DEUBAD independence.
The original PMID:16606617 paper was read through its primary-paper copy. Targeted Plasmids/Cloning, yeast/GST/IP/reporter Methods, Results and Figure 4–7 text captions were inspected; figure pixels, the full supplementary record and the entire paper were not. Figure 5A uses human ASXL1 residues 1096–1541 with human ESR1, mouse RXRalpha and chicken TRbeta. The GST assay uses in-vitro-translated receptors; it is not a purified binary affinity measurement. Figure 5B and cellular coactivation assays use mouse ASXL1 expression. The abstract-only cache and its availability flags are unchanged.
Official partner entries are P03372, human ESR1, P28700, mouse Rxra and P68306, chicken THRB; NCBI chicken THRB, Gene396431 corroborates the latter. All three partners already occur in the shipped UniProt INTERACTION block. Its RP/CC summary is not an exhaustive interaction list. The historical Xeno discussion above describes a corrected interpretation of a xenogeneic-interaction flag, not a Xenopus species assignment for these accessions.
PMID:20436459 directly reconstitutes human BAP1/ASXL1 and measures ASXL1-dependent nucleosomal H2A deubiquitination. PMID:22897849 instead supplies ASXL1–BAP1 co-IP and distinct HOXA responses to perturbation; its association record remains non-core. Evidence for a partner elsewhere does not establish the precise assay of an unread source-specific record. PMID:19615732 general Methods were read but its target Table S3 edge was not. For PMID:32296183 the HuRI target pair/construct validation remains unread despite article availability. PMID:33961781 and PMID:35512704 caches contain partial Introduction/Discussion text; target-specific Results/Methods and pair matrices remain unavailable to this assessment. Their existing true full-text flags are preserved. These six uncertain interaction rows are not rejected as false. The separate policy preference to avoid generic binding does not itself satisfy the user definition of REMOVE as likely incorrect biology.
Mouse Asxl1 disruption produces structural developmental abnormalities; human ASXL1 variants cause a congenital-malformation syndrome.
The independent developmental source consultation read the complete official PMID:21706002 abstract on human Bohring-Opitz syndrome, the complete official PMID:19833123 abstract and displayed Figure 2 caption on mouse Asxl1 skeletal homeotic transformations, and the complete official PMID:24218140 abstract with selected original Results/displayed captions on mouse ocular, craniofacial and skeletal defects. These observations are attributed to that bounded independent reading; this followup does not claim a new whole-paper, figure-pixel or PAINT-tree reconstruction. The three papers are not local publication caches and receive no invented cache-availability flags. They support retention of the existing non-core developmental context, not a new human organ-morphogenesis mechanism. The leukemia paper PMID:22897849 is no longer the sole attachment for that positive developmental claim.
Official MGI identifies MGI:2684063 as mouse Asxl1 and MGI:1922552 as mouse Asxl2; Q76L83 is human ASXL2. This corrects the earlier plural ASXL2-source wording for the PPAR IBA row. All original WITH/FROM values remain unchanged. Authored IBA provenance now uses the actual PTN as its proximate ancestral assertion and describes extant evidence separately. Different PTN identifiers do not establish topology or lineage-specific gain/loss without the tree. Human Q8IXJ9 as its own experimental descendant source is legitimate, not circular.
The DNA-binding row remains ACCEPT: human DEUBAD/nucleosomal DNA contact and a separate HARE-HTH oligonucleotide experiment support broad binding while leaving the PHD mapping unresolved. Neither supports assigning DNA recognition to the PHD region. No additional DNA-binding term or biological process is proposed.