The source-evidence sidecar records the exact incoming contexts, product identities, manual decisions and the provenance of this update. Its raw contexts retain the two distinct same-key cDNA annotations, all four grouped extensions, and the source's unsuffixed product identifiers.
Evidence boundaries are described in the manual synthesis. Primary sources, database assertions and reviewer interpretations are distinguished. A cached abstract remains an abstract even when selected original sections or a supplement were read elsewhere. Native publication titles and cache availability flags are taken from the actual source files.
The strongest biochemical support concerns the active lysosomal product. The separate extracellular-matrix-binding proposal is restricted to P16278-2. No new process term is inferred from substrate accumulation alone, receptor association, endothelial staining, or release as granule/vesicle cargo. Matrix binding and galactoside binding have no ancestor/descendant relation in the checked ontology. No matching GLB1 activity was found in the inspected local GO-CAM index or cached models; this bounded absence is not evidence of a curation omission.
The selected original material resolves the human enzyme's recombinant dimer, S-Gal ligand association, the NGX6 citation-target concern, HuRI partner identities, human glycan measurements and the target-specific urinary-vesicle detections. Rat donor originals remain unavailable beyond authentic abstracts. The resulting uncertainty applies to the exact binding/process claims, not to whether D3ZUM4 is rat Glb1.
The prior review, 64 historical quote associations, 59-row prior GOA snapshot, original notes prefix and six historical session records are retained in the provenance closure. Their retention does not endorse every former interpretation. The update uses compact representative quotations and explicitly supersedes the earlier private raw35/raw60 proposals after additional term and supplement checks.
Native acquisition and transfer passed schema, ontology-term, reference and strict GOA checks. Review follow-up harmonized the repeated-term decisions and direct citation chains; validation now reports no warnings and computes COMPLETE. Five explicit UNDECIDED judgments remain scientific access/scope limits. The initial DRAFT receipt and its two advisories remain preserved in the history and source-evidence sidecar. The current EDIT record and tracked HTML accompany the follow-up.