Skip to content

Add evidence-backed toxicyst discharge phenotype - #1837

Merged
realmarcin merged 1 commit into
mainfrom
codex/add-toxicyst-discharge
Oct 8, 2026
Merged

realmarcin merged 1 commit into
mainfrom
codex/add-toxicyst-discharge

Conversation

@realmarcin

Copy link
Copy Markdown
Contributor

Summary

  • Add traitmech:000685, toxicyst discharge, as a PROPOSED PHYSIOLOGY class.
  • Three primary DOIs and directly read Abstract/Summary snippets; controlled discharge is distinguished from injury, organelle possession and sequence predictions.
  • Use phenotype METPO:1000059; the exocytosis hierarchy remains explicitly unresolved. No unverified canonical taxa, xrefs or protein mechanisms.
  • Reserve v561 and METPO:1063800-1063899 with one class row. Add guarded writer/tests, append-only history and regenerated published artifacts.

Evidence

Definition authority: https://doi.org/10.1111/jeu.12106 (PMID:24512001), living Coleps discharge; abstract-only access.

Experimental triggering: https://doi.org/10.1007/BF01279249, publisher Summary and author-upload full-text transcript https://www.researchgate.net/publication/227062328_Ca2_triggers_toxicyst_discharge_in_Didinium_nasutum. Figures were not visually inspected.

Mechanism limits: https://doi.org/10.1186/s12915-024-01904-2 (PMID:38715037; PMC11077807). Figure 1 A-C are drawings, D-F fixed-cell TEM; genomic associations are not discharge perturbations. Supplements not used.

Stable restricted postprint location: https://hdl.handle.net/11393/192287; no full-text access claimed.

Resolver outcomes: two VERIFIED, one UNRESOLVED (1999 DOI); manual direct publisher check retained separately. No baseline exceptions.

Novelty and preservation

Ignored-and-hidden searches cover 20 existing worktrees, whole CommunityMech and complete paginated open heads including drafts. No toxicyst terminology/source matches; all 110 candidate-allocation hits are unrelated numbers. Main/open reservations rechecked before push.

All 1,079 prior YAML records, history, proposals and embedding artifacts are byte-preserved. 1,078 old pages have footer-only changes; phenotype gains the new child. Priority parent child count changes 177 to 178. Corpus total is 1,080. Both configured embedding sources are absent, so embeddings were not regenerated.

Validation

LinkML, strict validation, proposal verification and ROBOT/ELK pass. Parsed RDF confirms w3id hierarchy 1063800 -> 1000059 -> 1000188, including parent label/definition; RDF triples: 15 classes, 12,628 merged, 12,632 reasoned. Grounding/Biolink, graph/taxon/domain, snippet, history/product audits and Ruff pass.

The 20 new writer tests pass. Desktop/mobile Chrome at 1440/390px passes identity, evidence, history, hierarchy/browse navigation, dashboard image/count, no overflow or JS errors; all six screenshots inspected.

Ordinary staged whitespace check flags only the required three trailing directive cells on proposal TSV line 2; exact 11-column headers match pinned upstream and narrowly scoped checks pass.

Committed-tree just qc, committed history audit against origin/main, and PR sanity pass. The focused suite passed 65 tests in 363.06 seconds; full pytest passed all 3,621 tests in 705.97 seconds. All local validation processes completed successfully.

Review and merge

Author adversarial review challenged identity, deduplication, parent semantics, source section/quotes, injury and genomic confounds, proposal allocation and artifact preservation. No actionable new curation defect found. Two scientific discussions remain explicitly OPEN in the record.

The add-trait skill already covers the source/sequence/figure distinctions used here; no additional guidance change is needed for this record.

Copilot will be requested once. Latest manual Claude run 37410621810 and Shepherd run 37410625895 remain quota-exhausted, as documented on #742; no redispatch or paid fallback. Missing independent review is not an approval. Follow the required-check/native-queue path, verify landed-tree integrity, then delete only this feature branch/worktree.

@realmarcin

Copy link
Copy Markdown
Contributor Author

Author adversarial review of exact head d3d5ea8 completed. Challenged: toxicyst-vs-other-extrusome identity; actual release vs possession, injury and predation; phenotype vs fusion-defined exocytosis parent; direct scientific Abstract/Summary provenance and short quotations; restricted/full-text/figure access limits; sequence-based mechanism overreach; complete open-head/local-worktree reservations; and generated artifact scope. No actionable new curation defect found. Both scientific discussions remain OPEN; no independent review or human promotion of PROPOSED status is claimed.

All 1,106 remote changed-file blobs match the reviewed local commit. All 1,079 existing YAML records, prior history/proposals and embeddings are preserved; old pages have only footer totals plus the phenotype child link. Local QC/history/products/schema/proposal/ROBOT checks and 3,621 tests pass (focused 65). Six desktop/mobile screenshots inspected. Required TSV trailing directive cells are the sole ordinary whitespace finding, with exact-width and scoped checks passing.

Copilot was requested once; response contained an empty requested-reviewers list, not a submitted review. Claude/Shepherd latest manual attempts remain quota-limited with no agent diff review; availability receipt: #742 (comment). No retries or paid fallback. Required CI and latest feedback must be checked before native queue admission.

@realmarcin

Copy link
Copy Markdown
Contributor Author

Pre-merge exact-head audit: all eight required checks from integration 15368 are completed/success on d3d5ea8. PR CI pytest passed all 3,621 tests in 307.99 seconds; local full suite passed the same count. Fresh paginated formal/inline feedback is empty; conversation contains only the author review receipt and no actionable finding. Independent review remains unavailable, not approved, as tracked on #742. Current rules require zero submitted approvals and native SQUASH/ALLGREEN merge queue. Proceeding under the user explicit commit/review/merge/cleanup instruction via normal queue admission, with no admin bypass; landed-tree verification precedes branch deletion.

@realmarcin
realmarcin added this pull request to the merge queue Oct 8, 2026
Merged via the queue into main with commit 9c7ae6a Oct 8, 2026
8 checks passed
@realmarcin
realmarcin deleted the codex/add-toxicyst-discharge branch October 9, 2026 07:33
@realmarcin

Copy link
Copy Markdown
Contributor Author

Completed: merged 2026-10-08T19:06:30Z as 9c7ae6a through the native queue. All eight PR-head and eight merge-group checks succeeded. All nine post-push workflows and Pages deployment 37829449809 succeeded. Local tests: 3,621 pass; focused tests: 65 pass; PR CI: 3,621 pass.

Maintained verify_merge_integrity against actual parent 3db67c7 reports complete=true, verified=1, errors=[], mismatch=false; expected and actual tree both 1947389c9b4237df410282d5ba9b0a729be312e0. This verification preceded remote branch deletion.

On resumption, formal and inline reviews were still empty and conversation feedback contained no new findings. Author adversarial review found no actionable curation defect; independent reviewer availability remains a limitation tracked on #742, not an approval.

Remote branch deleted with an exact-head lease; remote absence confirmed by ls-remote exit 2. Local/tracking branch and owned worktree removed. Main fast-forwarded cleanly to e7be6f1, preserving the subsequently merged #1838. All 21 other worktrees and the primary venv are preserved.

Live page https://culturebotai.github.io/TraitMech/pages/traits/physiology/toxicyst_discharge.html exactly matches the merged artifact: 12,394 bytes, SHA256 6f482e977bb870afe592c3a47bf0db1d439e525fad532617a29f80bf7704c95a. Two scientific discussions and the upstream v561 proposal remain open. Discovery is not exhausted.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant