## ERO1A
- **UniProt:** Q96HE7 · **batch:** proteostasis-batch-2026-06-07b · **review status:** COMPLETE
- **PN placement:** `ER proteostasis|Folding enzyme|Protein disulfide isomerases|Protein disulfide isomerase reoxidation`. **PN-node mapping:** type `PDI reoxidation`=mapped→**GO:0016971 flavin-dependent sulfhydryl oxidase activity** (corrected in batch-5 after gene-level review); parent group `Protein disulfide isomerases`=mapped→**GO:0003756 protein disulfide isomerase activity** (new_to_goa); class/branch=no_mapping.
- **Consistency:** Notes, review YAML, and the *corrected* PN type-node mapping are consistent: ERO1A is an FAD-dependent sulfhydryl OXIDASE that reoxidizes PDI, not an isomerase. Review core MF = GO:0016971 (EXP PMID:11707400/29858230, ACCEPT) and it explicitly **MARK_AS_OVER_ANNOTATED** the GO:0015035 protein-disulfide reductase rows (wrong directionality). The dossier rationale documents this correction explicitly. **Internal PN tension:** the parent `Protein disulfide isomerases` group still maps GO:0003756 and the projected-annotations list still projects GO:0003756 to ERO1A as new_to_goa — which contradicts the corrected type node and the review (which never asserts isomerase activity for this oxidase).
- **PN story / NEW pressure:** GO:0016971 (OLS-verified) is already in GOA (EXP) and ACCEPTed — captured, no NEW pressure. GO:0003756 (verified real) should NOT be added to ERO1A: it is biologically wrong for an oxidase (the review asserts oxidase, not isomerase; cf. the reductase over-annotation it rejects). Conclusion on the group projection: **over-reaches** for ERO1A.
- **Mapping strategy:** The type-level correction (→GO:0016971) is exactly right and the model fix to follow. The problem is purely inheritance: the group node `Protein disulfide isomerases` lumps catalytic PDIs (P4HB), non-catalytic members (ERP27), and the EROs (oxidases), so its GO:0003756 projection mislabels the oxidase children. Note GO:0016971 is itself a GO child of GO:0015035 reductase, so the over-annotation the review rejects is the *parent* term, not the accepted child — the directionality argument (not the hierarchy) is what justifies the review's split.
- **Evidence alignment:** PN dossier lists no reference titles. Review oxidase evidence (PMID:11707400, 29858230, 10671517, GOA-anchored to GO:0016971/GO:0016972) is reviewer-supplied; the ERO1A↔P4HB redox-relay interaction (PMID:20802462) cross-links to the P4HB review (where it appears as ERO1A protein binding). Alignment is by shared biology and the cross-referenced relay partner.
- **Verdict:** Consistent at the (corrected) gene/type level; GO:0016971 validated and captured. The lingering **group-level GO:0003756 projection over-reaches** and should not land on ERO1A. **Recommended edits:** [MAP] suppress/override the GO:0003756 projection for ERO1A (and ERO1B) so the PDI-reoxidation oxidase children inherit only GO:0016971, not the parent isomerase term (matches the already-corrected type node and the review's oxidase-vs-isomerase distinction).