G6PC1 is the ER-membrane glucose-6-phosphatase catalytic subunit. The review distinguishes its hydrolytic core, condition-dependent phosphotransfer measured in the mouse ortholog, and secondary nutritional, hormonal and drug responses. The 357-residue canonical protein and 176-residue alternative product remain distinct; engineered structural constructs are not natural isoforms.
The current judgments and manual reference assessments are in G6PC1-ai-review.yaml. The companion evidence archive retains original source contexts, source-to-inherited joins, product sequences, actual phylogeny, interaction-table provenance and quotation history. The private full proposal preserves the authored decisions; the compact public projection does not repeat them.
The earlier human cloning, disease-variant and topology sources were read as authentic complete abstracts unless a selected-original scope is named above. An abstract foregrounding SLC37A4 or G6PC3 is not evidence of a wrong-gene curation error. Reactome disease variants are kept distinct from wild type, and the gene in a transcriptional event is distinct from the ER-membrane protein output.
All four IBA origins are actual PTHR12591 AN0/PTN000971890. Human AN15/PTN002488968 has the exact current 357-residue sequence and lies below that origin. Literal donor sets, including self-donors, agree with the retained IBD records. Three seed-order differences in the native family table are preserved as serialization differences. No source-count argument or pairwise-similarity substitute is used.
The organelle policy retains supported ER-root assertions alongside ER membrane. Positive generic binding is retained non-core under the standing campaign instructions.
The inherited source export has 40 contexts but the stored review has 39 because two partner contexts were combined. Current originals remain literal; 16 date changes are source-version differences. All 42 inherited quotation associations and the complete old review are retained as historical provenance. Only two short inherited core excerpts remain active, each matched to its immutable source. Historical wording is not current evidence merely because its text matches a source. The complete old notes, including obsolete rationale and tool-markup text, remain inert in the journal’s historical block.
Genuine native acquisition used fetch-gene --no-seed --force: only the target UniProt and GOA files were refreshed, and the inherited 39-row review remained untouched. All 40 actual source/export/parser contexts now have unique joins, including the two separate Haenig partner contexts. All 28 reference titles match their native records or unchanged GO_REF descriptions. Both retained core excerpts are exact native-cache substrings. The accepted transfer is now applied. Native EDIT history, target default and strict GOA validation, status, private rendering, scoped whitespace and source conservation passed. Status is DRAFT with three disclosed advisories: two supported generic-binding rows and the unused inherited research file. The initial transfer was published as PR #5833; its review prompted the follow-up below.
The source closure contains 56 objects: 21 existing immutable sources and 35 missing sources acquired through native functions (30 publications and five Reactome records). Full native text is available for PMID16625196, PMID27366200 and PMID39847333. The latter two were acquired during this stage and had already been read in selected original form for the accepted scientific proposal. Other publication records are abstract-only; their reference notes retain the actual reading limits. Failed full-text fallbacks for PMID7573034, PMID8182131 and PMID8733042 do not count as full text. Native availability does not expand a prior manual reading claim.
The genuine UniProt record also required Reactome R-HSA-9940951, a pathway for NuRD interactions with transcription factors. This additional cache concerns regulation of the G6PC locus; it supplies no nuclear role for the G6PC1 protein and no new annotation or reference assessment. Its absent-path ownership check preceded acquisition. The compact evidence file retains every source identity and access flag, all current original contexts, literal donor order, the complete inherited review, and all 42 historical quote associations.
PMID:39847333(https://pmc.ncbi.nlm.nih.gov/articles/PMC11789071/) resolves contacts with the phosphate moiety of G6P in an engineered H176A substrate complex. Its apo Glu110 phosphate-surrogate interpretation does not demonstrate a separate inorganic-phosphate ligand. The phosphate-ion row remains unresolved on that ligand distinction, not because the paper is unavailable.
PMID:14988444(https://pubmed.ncbi.nlm.nih.gov/14988444/) is available as the authentic abstract. Its rat glucose/fructose infusions under pancreatic clamps give positive carbohydrate-response evidence, with condition-dependent hepatic G6pc expression at equal total sugar load. The literal donor's separate PMID:26883980 experiment remains incompletely resolved. The non-core action uses the independent positive infusion evidence without presenting it as a human assay.
The positive mouse phosphotransferase result in PMID:8407995 is preserved. Its physiological use is unresolved, so the action follows the measured-activity policy: MARK_AS_OVER_ANNOTATED. The assay is not denied and no noise-level comparison is asserted. The two already accepted glycogen- and glucose-6-phosphate-metabolism terms now appear in the existing catalytic core; no new GOA assertion is added.
The ER-membrane protein output is visible in the official external R-HSA-9944724 participant display. The immutable cached summary contains the transcriptional repression description but omits the participant compartment. The retained original external response establishes the reading scope; human G6PC1 structure independently supports the same protein location.
The short natural product's verified sequence lacks canonical His176 and much of the C-terminal membrane architecture. This is a sequence-based limitation on extrapolating the established mechanism, not an experiment proving that the product has no biochemical activity. Historical notes, markup, source objects, both active quotations and all 42 inherited quotation associations remain unchanged. The follow-up is applied. Target default validation and strict matching of all 40 GOA contexts passed. Status remains DRAFT with three disclosed advisories: two supported generic-binding rows and the unused inherited research file. Native EDIT history, private rendering, scoped whitespace, all 56 source identities and retained-state conservation passed. The pkg_resources deprecation message is retained separately. No source or ontology cache was changed. Amended publication and remote review remain pending.