Gamma is not merely consumed during clotting. Human gamma-containing intermediates assemble into fibrinogen (PMID:8910396), and gamma-derived polymerization-site experiments plus the a-pocket structure demonstrate molecular contacts used in fibrin assembly (PMID:6451630, Results/Figures 3–4/6; PMID:9207064, pocket Results). The structural core therefore has direct participation evidence. It does not assign the thrombin cleavage or factor-XIII crosslinking reaction to gamma. The broader structural-molecule term also covers assembly of the soluble complex; matrix-specific annotations are retained without a duplicate core based only on proteomic classification.
Human gamma preparations support platelet aggregation (PMID:6281794, Tables 1–2/Figure 1), but the isolated chains form multimers. The assays do not compare natural gamma-A and gamma-B or establish a soluble monomer as a platelet agonist. Official product identities are retained exactly: displayed gamma-B/gamma-prime is 453 residues; the alternative gamma-A C terminus yields 437 residues. Historical mature-chain numbering, precursor numbering, experimental fragments and pathological splice constructs must remain distinct.
Gamma117–133 recognition of ICAM-1 is supported by the positive PMID:7822297 abstract and independently read adhesion/survival experiments in PMID:10903502. This helps interpret receptor-triggered effects of intact fibrinogen. Human fibrinogen or multichain fragment D in rat vascular preparations (PMID:15739255) and rat heart endothelial cells (PMID:19193866) support contextual constriction, ERK and secretory responses. These are not demonstrations of isolated gamma hormone activity, gamma kinase activity or a gamma vesicle-fusion mechanism.
Fibrin-chain surface labeling/calcium-dependent retention is real in human platelets (PMID:6777381, Figure 3/Results). Granule cargo wording is not a gamma-specific intragranular image. Exact human tissue proteomic joins support ECM association; a urinary fraction contains a low-count FGG identification. Neither enrichment nor database protein naming supplies new mechanical, topology or splice-specific claims.
The fibrinolysis boundary remains open. The accessible PMID:16846481 abstract concerns an alpha-chain variant with normal plasminogen activation and delayed lysis; it does not settle gamma-specific template/cofactor work. Alpha-C oligomer results cannot be transferred to gamma. Genuine ARBA matches retain this uncertainty rather than being called impossible or erroneous. Other unavailable source-specific target tables and assays also remain UNDECIDED; positive generic complex/location evidence is distinguished from an unverified specimen or mechanistic claim.
The source guide records actual read scope and public locators. The source-provenance file preserves raw/TSV/seed joins, products, phylogeny, rules and component routes. One canonical review remains the decision authority. The configured research command was genuinely attempted and failed without producing a report; no failure cause was inferred and no provider report was fabricated. Manual primary-source work supplies this review. No new publication quotations or NEW annotation rows were added; source caches are immutable.
Validation retains nine explained advisories: three supported bare-binding rows follow the campaign policy, while six same-term checks span different source-access or assay scopes. Unresolved specimen/target joins are not converted into positive results because another source supports the generic term. The pathological truncation context remains non-core alongside the demonstrated structural role. All annotations have decisions, but the validation advisories keep the computed status DRAFT.
The initial review conflated source-table uncertainty with the generic claim in several places. Gamma polymerization and clot formation now receive ACCEPT from independent fragment/pocket evidence, without attributing the FGA Leu9Pro mechanism or directly measured protein truncation to the clinical papers. Generic exosome association remains non-core, supported by the exact low-count urinary-fraction identification; this does not verify the other paper's prostatic-secretion target row or intravesicular topology.
Five other raw records have real tissue/cell extensions omitted by seeding. Their generic functions or locations are supported, but the complete extended assertions remain UNDECIDED. The original predicates/identifiers are restored with official OLS labels and their exact raw objects preserved publicly. This source-metadata repair makes the different claim scope explicit without new annotations. Unknown gamma cofactor work and distinct unverified partners remain open.
The PMID:10788510 abstract supplies a verified adhesion-protein partner class as well as gamma binding. Its two existing binding annotations are refined to cell-adhesion-molecule binding; this does not turn the host protein into a bacterial adhesin. Affinity-capture association without that refinement remains supported generic binding, following the explicit campaign instruction. Eleven identical reference-only citations were removed. No new annotation, publication quotation, natural product or core function was introduced.
Follow-up validation passes with four explained advisories: one supported generic-binding row under the campaign instruction, and three same-term warnings whose differing specimen extensions are now explicit. The computed status remains DRAFT.