{ "version": "20260604_core_validation_v45", "generated_at": "2026-06-11T14:12:46+00:00", "go_no_go": { "status": "go_after_author_metadata_and_final_pdf_qa", "summary": "The resource is technically usable, curator-audited, DOI-backed, and live at the public HTTPS portal URL." }, "submission_snapshot": { "verified_release_evidence": 737, "toxicity_release": 626, "offtarget_release": 111, "benchmark_split_rows": 344, "candidate_records": 41114, "source_documents": 36245, "case_workflows": 5, "core_field_completeness_pct": 100.0, "p0_core_oligo_field_rows": 344, "independent_validation_reviewed_rows": 0, "independent_validation_sample_rows": 500 }, "adoption_status": { "external_users": "not_claimed_predeployment", "citation_status": "not_claimed_prepublication", "public_usage_analytics": "ready_to_collect_after_public_https_deployment", "evidence_to_collect": [ "server access logs and download counts", "public issue/correction submissions", "tutorial notebook reuse", "early-user feedback from oligonucleotide safety and ML groups" ] }, "public_release_blockers": [ { "item": "Stable public HTTPS URL", "status": "complete", "owner_action": "Live URL verified on 2026-06-11: https://oligovigil.pages.dev." }, { "item": "Archived data and benchmark DOI", "status": "complete", "owner_action": "Use archived DOI 10.5281/zenodo.20633779; keep release files and checksums unchanged for v1.0.1." }, { "item": "Public contact and maintenance page", "status": "pending", "owner_action": "Expose maintainer contact, update policy, and issue/correction route on the public site." }, { "item": "External usage evidence", "status": "not_yet_available", "owner_action": "Collect access logs, download counts, GitHub/issue activity, and early-user feedback after deployment." } ], "editor_questions": [ { "question": "Why will people use this?", "current_answer": "It unifies verified oligonucleotide safety/off-target evidence, provenance, curation audit, downloads, and benchmark splits in one no-login resource.", "risk": "Usage evidence cannot be claimed before public deployment.", "mitigation": "Use task-oriented workflows, DOI-ready downloads, and post-deployment analytics rather than invented adoption claims." }, { "question": "What is new versus RNA/drug databases?", "current_answer": "The release is safety-first and curator-audited, with exact source locations, toxicity/off-target separation, candidate-release boundaries, and reusable benchmark splits.", "risk": "Generic RNA/drug resources may look broader.", "mitigation": "Lead with closest-work matrix and the safety/provenance/benchmark combination rather than raw database breadth." }, { "question": "Can records be trusted?", "current_answer": "737 release rows are curator-verified accepted records with audit trails.", "risk": "Sparse sequence, modification, and dose fields can make the resource look like a literature index.", "mitigation": "Expose the P0/P1/P2 core oligo field packet and avoid complete sequence/modification/dose claims until source-verified." }, { "question": "Is the curation error rate measured?", "current_answer": "Inter-curator agreement is measured: a blinded second curator (HY) re-scored a 100-row mixed inter-rater sample (KAPPA-2), giving Cohen κ_binary = 0.42 (moderate, drop-abstain, n=92) / 0.34 (fair, collapse-abstain, n=100), raw agreement 66% (66/100). A separate 500-row independent second-review packet (release accept rows plus rejected candidate controls) underwrites the release-row false-accept / false-reject error rates.", "risk": "Cohen kappa is claimed; the release-row false-accept / false-reject error-rate estimates remain pending until the 500-row independent-check packet is complete and adjudicated (A10 third-adjudicator pass still pending).", "mitigation": "Use /api/independent_validation and independent_curation_validation_template.csv as the validation gate." }, { "question": "Can ML groups cite it?", "current_answer": "344 Grade A/B rows are exposed through fixed reference splits and task cards.", "risk": "Baseline model results remain pending.", "mitigation": "Ship reference splits now and add transparent baseline result tables after public freeze." }, { "question": "Will it stay available?", "current_answer": "The portal supports no-login access, CSV/ZIP downloads, manifests, maintenance commitments, and a DOI-backed archive (10.5281/zenodo.20633779).", "risk": "Long-term usage evidence still needs to be collected after public launch.", "mitigation": "Track public access logs, download counts, issue/correction submissions, and early-user feedback after deployment." } ], "case_study_cards": [ { "title": "GalNAc-siRNA liver safety profile", "audience": "oligonucleotide drug discovery teams", "release_records": 22, "primary_endpoint": "/api/modification_profile?term=galnac", "benchmark_task": "toxicity_safety_v0_1" }, { "title": "ASO/gapmer hepatotoxicity review", "audience": "ASO chemistry and safety groups", "release_records": 341, "primary_endpoint": "/api/evidence_records?domain=toxicity&q=hepatotoxicity", "benchmark_task": "toxicity_safety_v0_1" }, { "title": "Renal and thrombocytopenia safety scan", "audience": "preclinical safety reviewers", "release_records": 51, "primary_endpoint": "/api/search?q=renal", "benchmark_task": "toxicity_safety_v0_1" }, { "title": "siRNA seed/off-target transcriptome evidence", "audience": "RNAi design and transcriptomics groups", "release_records": 44, "primary_endpoint": "/api/evidence_records?domain=offtarget&q=seed", "benchmark_task": "offtarget_safety_v0_1" }, { "title": "Benchmark reuse with fixed reference splits", "audience": "ML and computational biology groups", "release_records": 737, "primary_endpoint": "/api/benchmark", "benchmark_task": "toxicity_safety_v0_1 / offtarget_safety_v0_1" } ], "quality_checks": [ { "check": "no_login_access", "status": "pass", "evidence": "All API and static portal endpoints are unauthenticated." }, { "check": "bulk_download", "status": "pass", "evidence": "CSV tables, populated evidence_release.csv, benchmark_reference_splits.csv, manifests, and all_tables.zip are exposed." }, { "check": "candidate_release_separation", "status": "pass", "evidence": "Candidate annotations remain separate from release evidence tables." }, { "check": "release_promotion_gate", "status": "pass", "evidence": "Release evidence explorer exposes only curator_verified, accepted A/B/C records; abstract-level generated promotions are excluded." }, { "check": "human_verified_release", "status": "pass", "evidence": "737 of 737 release rows carry a human curator-verified accept audit (independent re-curation of 2003 v1 machine pre-curated candidates)." }, { "check": "crispr_exclusion", "status": "pass", "evidence": "CRISPR guide RNA remains excluded from core molecule records." }, { "check": "stable_public_url", "status": "pass", "evidence": "Cloudflare Pages URL https://oligovigil.pages.dev resolved and passed live endpoint checks on 2026-06-11." } ], "field_completeness_summary": { "core_required_avg_pct": 100.0, "records_with_any_sequence": 0, "records_with_any_chemistry_or_delivery": 0, "sequence_completion_status": "gap", "chemistry_completion_status": "gap", "action_note": "Core provenance fields support citation now; sequence and chemistry fields remain the highest-value expansion path before public submission." }, "recommended_next_actions": [ "Keep https://oligovigil.pages.dev live and repeat smoke/final checks before submission.", "Keep the frozen data bundle tied to DOI 10.5281/zenodo.20633779 and do not mutate the v1.0.1 release files.", "Add baseline result table after DOI freeze without changing reference splits.", "Expand exact sequence/modification curation to make sequence search a primary use case." ] }