The official human FMR1 record is HGNC:3775 / UniProt Q06787. Genuine fetch and seeding produced 250 annotations from 253 original QuickGO contexts and 11 natural products. Existing source files were hydrated from exact main before acquisition. All 85 required normal sources are now present, with 18 original main files preserved byte-for-byte. No new workspace, environment, clone or bulk leaf archive was created.
The review combines four manually assessed groups: 51 biological-process contexts, 61 other molecular-function contexts, 43 generic-binding contexts and 98 cellular-component contexts. All were independently cross-reviewed before integration. The final stored counts are 81 ACCEPT, 138 KEEP_AS_NON_CORE, 9 MODIFY, 6 TOO_GENERAL and 16 UNDECIDED. The source guide records the actual publication-access boundaries and the evidence file preserves original contexts, extensions, products, family joins and source hashes.
The two cores summarize mRNA recognition/regulation and condensate scaffolding. Target-dependent translation can be positive or negative. The action-potential, channel, viral, developmental and restricted nuclear/chromatin findings remain contextual. The AMPA-internalization row is refined to the experimentally supported negative direction while retaining its original source. The positive-filopodium assertion is generalized to regulation; positive spine development and the presynaptic-translation IBA remain unresolved at their stated scope.
Eight original extension groups (12 conjunctive atoms) were restored through the supported schema slot. No original term, reference, evidence code, qualifier, partner or product was replaced. Three raw-context pairs legitimately project to one seed row each; their originals remain available. Historical isoform names are not mapped to accession suffixes by number, and fragment experiments are not applied to every product.
The standing user instruction in CLINGEN_MENDELIAN takes precedence over the default generic-binding skill guidance and quotation request. Supported generic interactions are retained as non-core, specific binding classes are used where justified, and inaccessible or insufficiently resolved contexts remain undecided. No new publication quotation or NEW annotation was added; validator defaults are unchanged. Research was manually source-based after known provider access failures, without inventing a provider report.
See the source guide for readable assay limits and publication links, and source evidence for exact provenance. Validation, rendering and final provenance outcomes will be appended after execution.
Full schema, term and best-practice validation passed after correcting the private extension prototype to the schema's CURIE-string predicate representation. All eight original conjunctions, 12 relation IDs and targets remain unchanged. The source evidence retains the predicate labels separately. The presynaptic-translation IBA now also records the unresolved mechanism in structured propagation metadata without disputing the verified ancestral-node placement.
Strict GOA matching, source/context conservation, product reconstruction, history validation and private rendering passed. The repository computes DRAFT with 37 advisories: 32 supported generic-binding rows retained under the standing user instruction, plus five same-term action differences. These distinguish source-specific mRNA/adaptor extension uncertainty, restricted cytoplasm refinements, unresolved poly(U)/poly(G) assays and source-specific self-association evidence. The differences do not assert that the verified interactions or the broader cellular locations are false. Validators were not changed. The separate environment deprecation warning is unrelated to curation status.
The final record preserves 250 seeded source contexts in order, all 11 natural products, 85 normal-source files (18 exact-main files), and the original 253-row projection. It contains two cores, 92 references and no new publication quotes or NEW annotations. All scientific partitions received independent review; final source-access clarifications distinguish the negative minigene assay from altered endogenous splice proportions in PMID:18653529 and identify the selected primary passages actually read for PMID:19097999, PMID:19005212 and PMID:18805096. No global test-suite execution or publication is claimed here.
The completed review requested explicit translation-activity coverage and clearer handling of two extended signaling assertions. Independent scientific consultation supported three distinct cores: mRNA recognition, bidirectional translation regulation and condensate scaffolding. The translation core uses GO:0045182 and retains both positive and negative processes, with RNA, construct and species boundaries.
The two PMID:31439799 rows remain UNDECIDED for their complete signaling-related conjunctions. Their base RNA-binding and condensate evidence is positive and accepted in other contexts; every original extension atom is preserved. The PMID:18664458 broad cytoplasm row changes from MODIFY to KEEP_AS_NON_CORE while retaining the actual TDRD3-induced stress-granule observation. Five central references gain bounded findings without supporting quotations.
Current totals are 81 ACCEPT, 139 KEEP_AS_NON_CORE, 16 UNDECIDED, 8 MODIFY and 6 TOO_GENERAL across 250 stored annotations. The original 253 contexts, 11 products, eight extension groups/12 atoms and 85 normal sources are unchanged. The standing user instruction on generic-binding decisions and reference-only support continues to apply; validators are unchanged. Validation and the exact final conservation result will be recorded after the checks complete.
Scoped validation passed for the exact revised YAML: schema and ontology terms, strict GOA matching, generated-history schema, report-only status, rendering and exact release/conservation replay. The computed status remains DRAFT with 37 advisories: 32 standing-policy generic-binding advisories and five source-specific same-term differences. The runtime also reports its existing pkg_resources deprecation warning. All 247 other reviews, original fields/order, 11 products, eight extension groups/12 atoms, 85 source hashes and the published history are unchanged. The prior independently reconstructed product-sequence proof is reused because the records are exact; no new sequence analysis or global test suite is claimed.