ERCC8 review notes

2026-10-24 — annotation review

All 77 imported contexts have been reviewed using independently checked database provenance and recovered primary evidence. The decisions are 48 ACCEPT, 14 KEEP_AS_NON_CORE, 11 MODIFY and four UNDECIDED. The unresolved assertions concern CUL4B-complex membership, CSA participation in its own ubiquitination, regulation of TC-NER and classical end joining. Supported generic binding remains non-core under campaign policy. The 396- and 205-residue products remain distinct. No NEW annotation or publication quotation was added.

Falcon and its fallback failed during the genuine research-provider attempt; no provider report was fabricated. Original-source access and precise reading limits are documented in the source guide. Core synthesis, final status and rendering follow this annotation checkpoint.

Core synthesis and final integration (2026-10-24)

The annotation checkpoint above remains preserved: all 77 reviews and source contexts, 43 references and two products are unchanged. The independently accepted description and one substrate-adaptor core now connect CSA's CUL4A-ligase assembly role to transcription-coupled nucleotide-excision repair and protein ubiquitination. The core retains full-length human CSA/mouse RBX1/pig polymerase reconstitution limits and distinguishes the separate 2020 Xenopus ligase experiment; equivalent activity of the 205-residue product is unestablished. Three biological questions and experiments address this short product, CSA participation in its own ubiquitination and classical end joining versus damage signaling. No NEW annotation or publication quotation is introduced. The accepted source guide is retained with explicit core evidence appended. Final computed status and validation outcomes follow below.

Final validation result (2026-10-24)

Normal validation, including GOA reconciliation and term checks, strict whole-record term validation, ontology component checks and rendering passed. The actual status manager computes DRAFT with five advisories, all for supported generic protein-binding annotations retained as non-core under campaign policy. No annotation action was changed to silence a warning. The only final metadata change is INITIALIZED to the computed DRAFT status. All 77 reviewed contexts, 43 references and both products remain exact; the 78 raw/TSV contexts retain their accepted mechanical collapse to 77 reviewed contexts. All 156 description, annotation and core narratives, plus 12 question/experiment fields, appear in the generated HTML. Delivery retains all original normal-source bytes and uses explicit ADD/REUSE comparison with pinned main, without shared-source overwrite.

Scientific feedback revision (2026-10-24)

Seven review objects receive explicit updates: chromatin replaces the nucleoplasm proposal for the 2020 chromatin-associated repair assembly; the conditional multi-insult redistribution response remains non-core with its positive scope stated; three operational nuclear-matrix assertions become non-core; the autoubiquitination uncertainty is made mechanistically explicit; and regulation of TC-NER is corrected laterally to independently demonstrated TC-NER participation. The adaptor core location becomes chromatin without adding an ancestral location or changing its molecular function, processes or complex.

All 77 imported source objects, 43 references and two products are unchanged, with no NEW annotation or publication quotation. The final annotation counts are 45 ACCEPT, 17 KEEP_AS_NON_CORE, 12 MODIFY and 3 UNDECIDED. Original source scopes remain distinct from independent evidence used to resolve the gene-level role. The existing notes above remain the historical journal. The source guide records the actual pages, Methods, figures and remaining mechanistic limits. The full accepted PMID:26620705 normal cache remains unchanged; no replacement by a shorter shared record is proposed.