## ARF1

- **UniProt:** P84077 · **batch:** proteostasis-batch-2026-06-03 (Falcon DR 2026-06-07) · **review status:** COMPLETE
- **PN placement:** 1 row, ER. `ER proteostasis|Protein transport|ER-Golgi trafficking|Retrograde transport|COPI coating and uncoating`. **PN-node mapping:** subtype=`mapped`→GO:0030126 COPI vesicle coat (more_specific); type (Retrograde transport)=`mapped`→GO:0006890 retrograde Golgi-to-ER transport (more_specific); group (ER-Golgi trafficking)=`no_mapping`; class (Protein transport)=`mapped`→GO:0015031 (entailed_by_goa_closure); branch=`no_mapping`.
- **Consistency:** Fully consistent across DR, notes, review YAML, and PN. The notes explicitly walk all three projected terms. No contradiction.
- **PN story / NEW pressure:** PN asserts retrograde Golgi-to-ER transport + COPI coat handling. Review ADDS GO:0006890 (verified real; `action: NEW`, TAS, Reactome:R-HSA-6811434) — defensible, matches PN type-node projection. GO:0015031 already captured (entailed via GO:0006886 IBA). **GO:0030126 COPI vesicle coat: PN subtype over-reaches** — ARF1 is a regulatory GTPase that recruits/releases coat, not a coatomer subunit; review correctly REJECTS adding it (mirrors the TOMM20/RAB7A "wrong-bucket" precedent — a CC membership term projected onto a regulator).
- **Mapping strategy:** ARF1 does not change the node mapping. The retrograde-transport (GO:0006890) and protein-transport (GO:0015031) projections are right. The COPI-coat subtype (GO:0030126) is sound for coatomer members but should not project to ARF1; review files this as a suggested question (distinguish structural coatomer from regulatory GTPases).
- **Evidence alignment:** PN row carried no PMIDs; review/notes anchor on PMID:8253837 (ARF GTP cycle), PMID:10102276 (ARF1-ARFGAP-coatomer), Reactome:R-HSA-6811434. Falcon DR added PMID:37914730 (cGAS-STING), 37400497 (FA/mito), 36269825 (AP-1:Arf1 coat) as statement-only, non-core. No divergence.
- **Verdict:** Consistent; GO:0006890 NEW is warranted and matches PN. PN GO:0030126 over-reaches for ARF1 (regulator vs coat subunit) and review correctly declines.
- **Recommended edits:** none to ARF1-ai-review.yaml. [MAP] flag PN "COPI coating and uncoating" subtype to not project GO:0030126 (CC membership) to regulatory GTPases such as ARF1; restrict to coatomer subunits.