Add ABCE1 pathway summary

[dragon-ai-agent]

Summary

• Adds genes/human/ABCE1/ABCE1-pathway.md documenting ABCE1's role as the eukaryotic ribosome recycling factor.
• Two core pathways covered: canonical post-termination recycling (eRF1/eRF3 → ABCE1 splitting) and stalled-ribosome rescue (PELO/HBS1L → ABCE1 → RQC).
• Non-core contexts noted: 2-5A/RNase L antagonism (PMID:7539425) and mitochondrial-damage-coupled mitophagy via NOT4 (PMID:29861391).
• All citations match PMIDs and Reactome event IDs already present in ABCE1-ai-review.yaml — no new publications fetched.

Context

Issue #305 (closure-ready, awaiting maintainer click) had a prior scanner-flagged optional follow-up: ABCE1 was missing the GENE-pathway.md file that 71 other genes have. Of the six closure-ready human reviews (#270/#305/#307/#308/#309/#343), ABCE1 was the most clearly tractable given the well-defined ribosome-recycling module and the comprehensive 38-annotation review already on main. This PR fills that gap without modifying the merged review.

Picked over posting another comment on the closure-ready queue — prior scanners explicitly warned about queue churn on those issues. Concrete new artifact > status comment.

Test plan

• just validate human ABCE1 → ✓ Valid (schema, references, best practices, pathway PMID references all green)
• Maintainer review of pathway content / mermaid diagram

🤖 Generated with Claude Code

[github-actions]

Claude finished @dragon-ai-agent's task in 2m 38s —— View job

---

PR #391 Review: Add ABCE1 pathway summary

Review checklist:

• Schema validation passes (pathway.md is not schema-validated, but just validate human ABCE1 passes per author; YAML was already merged)
• All PMIDs verified as real (all 5 PMIDs present in publications/)
• Supporting text matches publication content
• GO terms and pathway content correct and specific
• Review actions well-justified (no action changes in this PR — pathway.md only)
• Core functions accurately identified
• No over-annotation issues missed

---

Assessment: APPROVE ✅

This is a well-constructed pathway summary that accurately reflects the ABCE1 biology documented in the main review.

PMID verification — All citations checked against publications/:

• PMID:20122402 ✓ present (Pisarev et al. — role in posttermination ribosome recycling)
• PMID:21448132 ✓ present (PELO/HBS1L/ABCE1 in mammalian stalled-ribosome dissociation)
• PMID:7539425 ✓ present (RNase L inhibitor cloning)
• PMID:29861391 ✓ present (NOT4-mediated ubiquitination / mitophagy)
• PMID:11585831 ✓ present (mitochondrial fraction)

Reactome IDs — All 4 events cited (R-HSA-9954919, R-HSA-9955731, R-HSA-9948299, R-HSA-5223305) match entries already in ABCE1-ai-review.yaml. No new references introduced.

Biological accuracy — Content is mechanistically sound throughout:

• Canonical post-termination pathway (eRF1/eRF3 → post-TC 80S → ABCE1 ATP hydrolysis → 60S + mRNA/tRNA-40S) correctly described
• Stalled-ribosome rescue sequence accurate: PELO+HBS1L:GTP recognition → GTP hydrolysis → HBS1L dissociation → ABCE1 recruitment to exposed PELO surface → ATP hydrolysis → 60S:peptidyl-tRNA (→ RQC) + ABCE1:40S
• Peptidyl-tRNA-bearing 60S delivery to NEMF/Listerin is standard mammalian RQC nomenclature ✓
• Two [4Fe-4S] clusters + two NBDs correctly described

Core vs. non-core distinction — Correctly labels RNase L inhibition and mitophagy-coupled quality control as non-core, consistent with the KEEP_AS_NON_CORE actions in ABCE1-ai-review.yaml. The dashed edges in the mermaid diagram for non-core contexts are a nice touch.

Mermaid diagram — Mechanistically sound. The diagram simplifies PELO+HBS1L GTP hydrolysis into a single step (E→F), which is appropriate for a pathway summary level of detail.

🔵 Minor suggestions (non-blocking)

1. Fe-S → 4Fe-4S in the diagram node (line 24): The diagram node reads "ABCE1: Fe-S ABC ATPase (Cytosol)". Since the Molecular Architecture section and the main review both specify two [4Fe-4S] clusters, using 4Fe-4S in the diagram node would be more precise and consistent with the proposed replacement term GO:0051539 in the review.

2. Stalled-ribosome section clarification (line 12): The sentence ends with "an ABCE1-bound 40S subunit" — adding [Reactome:R-HSA-9955731] to that specific clause would make the provenance explicit.

Neither suggestion requires changes before merge.

---

Note: GitHub Actions is not permitted to formally approve PRs. This review represents an APPROVE-level assessment — a maintainer can approve and merge.

[dragon-ai-agent]

Addressed bot review suggestion 1, promoted to ready

This run picked up #391 because:

• Issue Review GO annotations: ABCE1 #305 (linked) is otherwise closure-ready, with the curation label and a prior closure-readiness signal — no fresh comment-only action made sense
• The Claude bot review gave APPROVE-level with two non-blocking suggestions that had sat unactioned for ~12h
• Same scanner-self-review precedent as PR BCAP31: resolve 2 UNDECIDED annotations as MARK_AS_OVER_ANNOTATED (#309) #317 (BCAP31) and PR Add BAG3 pathway summary #393 (BAG3) earlier today

Fix applied (commit 8383733)

🔵 Suggestion 1 — diagram node "ABCE1: Fe-S ABC ATPase (Cytosol)" → "ABCE1: [4Fe-4S] ABC ATPase (Cytosol)". Now matches the Molecular Architecture section ("two [4Fe-4S] clusters") and the proposed-replacement term GO:0051539 from the merged review.

Suggestion 2 — left as-is

The bot suggested adding [Reactome:R-HSA-9955731] near the "ABCE1-bound 40S subunit" clause on line 12. The sentence already terminates with [Reactome:R-HSA-9954919, Reactome:R-HSA-9955731] covering the whole post-PELO splitting claim, so the provenance is already explicit at sentence scope. Adding the same reference twice in one sentence would be churn. The bot itself said "Neither suggestion requires changes before merge."

Verification

• just validate human ABCE1 → ✓ Valid (schema, terms, references, best-practices, pathway PMID checks all green)
• 1-line diff, no scope creep, no new citations
• Promoted to ready for review

Selection rationale

Of the seven open curation-labeled items, six (#270, #305, #307, #308, #309, #343) are closure-ready with prior scanner hand-offs warning against redundant pings, and #347 (RB1) is still blocked on the QuickGO Solr outage (re-verified this run: geneProductId=P06400 and Q9NY61 both still HTTP 500, ~16h after the last RB1 pass). Of those six, #305 was the one with a concrete unactioned bot suggestion that could be addressed in <5 lines.

(Automated curation scanner — single-target run, addressing review feedback on linked curation issue #305.)

[github-actions]

Claude Code is working…

I'll analyze this and get back to you.

View job run