Skip to content

Add evidence-backed fungal constricting-ring trap formation - #1824

Merged
realmarcin merged 1 commit into
mainfrom
codex/add-constricting-ring
Oct 8, 2026
Merged

realmarcin merged 1 commit into
mainfrom
codex/add-constricting-ring

Conversation

@realmarcin

Copy link
Copy Markdown
Contributor

Summary

Add traitmech:000677, fungal constricting-ring trap formation, as a
PROPOSED morphology class under phenotype METPO:1000059.

  • Two directly read positive primary studies support three-celled hyphal
    trapping structures, with formation kept separate from maturation/inflation.
  • Three short, contiguous evidence snippets: two positive studies and one
    terminology-boundary source, explicitly not a third positive trap study.
  • One culture/condition-qualified natural example: NCBITaxon:2743661,
    Drechslerella daliensis (nom. inval.), preserving live NCBI nomenclature.
  • METPO proposal v553 reserves METPO:1063000-1063099, using only 1063000.
    No upstream acceptance, equivalence mapping, synonym or causal graph claimed.
  • Append-only creation and correction histories, guarded validated writer,
    migration/idempotence/drift tests, regenerated pages/discussions/dashboards.

Closes #1823. The initial label was ambiguous with fungal cytokinesis;
the corrected label explicitly says trap, preserving the original history.

Sources

  • Chen et al. 2023,
    PMID:37968349 / PMC10651832: strongest identity/development evidence and
    mechanism lead. Relevant Results, Methods and actual Figure 1 inspected.
  • Zhang et al. 2022,
    PMID:36761641 / PMC9836436: natural-isolate morphology, Methods, taxon
    descriptions and actual Figure 1e inspected. Publication year is 2022.
  • Stachowiak et al. 2014,
    PMID:24914559 / PMC4137230: directly inspected abstract/Results establish
    the intracellular cytokinetic-ring terminology boundary only.

Two discussions remain OPEN: closer morphology hierarchy, and mechanism /
strain provenance / nomenclatural interpretation. No unsupported protein
mechanism or universal field-biological efficacy is inferred.

Scope and preservation

Ignored-and-hidden novelty/reservation searches cover 21 local worktrees,
CommunityMech and a fresh, fully paginated immutable open-PR snapshot (23 roots).
Pending heads remain #1802, #1782, #1476, #973 and draft #924 as documented in
the proposal. No competing exact record or reservation was found.

The artifact audit proves all 1,071 previous trait YAMLs, previous histories,
proposals and embedding artifacts are byte-preserved. 1,070 previous trait
pages differ only in corpus footers; phenotype additionally gains the new child.
Priority changes to existing rows are limited to phenotype children 171 -> 172.
The corpus is now 1,072 records. Protected curation is untouched.

Validation

Committed-tree just qc, history audit against origin/main, PR sanity and
the full pytest suite passed: 3,447 passed in 667.57 seconds.
The 22 writer tests, final LinkML/strict validation, proposal verifier,
ROBOT template/merge/ELK, history/products and Ruff have passed.
Parsed RDF verifies w3id METPO identities and the phenotype/quality ancestry:
15 class triples, 12,628 merged triples, 12,632 reasoned triples.

Snippet resolver: two VERIFIED abstract matches; the developmental quote is
NOT_IN_ABSTRACT and was verified directly in Results. All three source spans
are exact. Live NCBI audit: 754 examples, zero errors, 24 pre-existing warnings.

Final browser checks at 1440 and 390 px: identity/provenance, three snippets,
one example, two OPEN discussions, two history records, navigation, dashboard
counts and coverage image all verified; no overflow or JavaScript errors.
Desktop/mobile trait screenshots were visually inspected.

Ordinary cached whitespace checking flags only the required three trailing
tabs in the 11-column ROBOT directive header. Header widths and the canonical
upstream header were inspected; scoped whitespace checks pass without trimming
required cells. Embeddings were not regenerated because both configured source
files are absent at the explicitly checked paths.

Review

Author adversarial review found and repaired #1823 before publication; the
published diff will also be checked before queue admission. Copilot will be
requested once after opening. External Claude/Shepherd availability remains tracked on
#742; quota receipts are not reviews or independent approvals. No paid fallback.
The current add-trait skill already covers the demonstrated naming failure.

@realmarcin

Copy link
Copy Markdown
Contributor Author

Author adversarial review

Reviewed head: 5c3933e3595e00557e80a3b4d691281189b84439.
This is an author review, not independent approval.
The fully paginated GitHub file inventory contains 1,100 changed files; every
published blob matches this tested local commit, with head stability verified
before and after retrieval. Committed just qc, history and PR-sanity gates
passed; full pytest: 3,447 passed in 667.57 seconds.

Finding and repair

#1823: the draft's unqualified constricting-ring label also fitted intracellular
fungal cytokinesis. Direct primary abstract/Results inspection supports that
counterexample. The label now explicitly says trap; a third citation and
snippet document this terminology boundary, not an extra positive trap study.
The original event is preserved and the correction is appended. Writer tests
cover initial-to-final migration, history preservation and drift refusal.

Falsification checks

  • Identity/duplicates: ignored-and-hidden whole-root searches, pinned METPO RDF,
    fresh seed projection, related ring-shaped/mycelial traits and every pending
    PR's immutable curation files reveal no exact existing trap-formation record.
    The ring-shaped class concerns individual cells, not a multicellular trap.
  • Allocation: all local worktrees and fresh paginated open heads were checked
    for 000677, v553 and the entire 1063000-1063099 block. Returned foreign hits
    are unrelated numeric/sequence/test data or lexical matches, not reservations.
  • Evidence: directly read primary developmental Results/Methods/Figure 1 and
    natural-isolate taxon descriptions/Methods/Figure 1e support the differentia.
    Fully formed immature rings are not assumed immediately inflation-competent.
    All three snippets are contiguous source matches; resolver outcomes retained.
  • Example: soil origin and observed culture are specified. Live NCBI rank/name
    are retained including nom. inval.. Laboratory strain 29 remains evidence,
    not an unsupported natural exemplar. No universal field efficacy is inferred.
  • Hierarchy/equivalence: broad phenotype parent is honest pending a closer
    parent; no shifted xrefs, synonyms, SSSOM equivalence or disjointness added.
    Formation, maturation, inflation and cytokinesis mechanisms remain distinct.
  • Proposal: canonical 11-column headers, reserved ID, label/definition parity,
    and parsed w3id parent ancestry verified. No released METPO identity claimed.
  • Writer/history: fail-closed parent/target/template checks, prevalidation,
    dry-run default, idempotence and append-only provenance tested. Old records,
    history, proposals and embeddings remain byte-identical.
  • Products: browser checks at desktop/mobile verify identity provenance,
    evidence, example, discussions, history, links and dashboard counts without
    overflow or JS errors. Old-page differences are footer-only except phenotype's
    new child; existing priority rows change only that parent's child count.

No remaining actionable defect found in this review. The two explicit OPEN
scientific discussions are limitations, not claims of completed mechanism or
nomenclatural curation. The add-trait skill already covers the failure mode.

External review availability and the final validation/published-blob receipts
are recorded in the PR description and
#742 receipt.
Unavailable reviewers are not clean approvals.

@realmarcin
realmarcin added this pull request to the merge queue Oct 8, 2026
Merged via the queue into main with commit dc6dcf6 Oct 8, 2026
8 checks passed
@realmarcin
realmarcin deleted the codex/add-constricting-ring branch October 8, 2026 10:01
@realmarcin

Copy link
Copy Markdown
Contributor Author

Verified landing and cleanup

PR #1824 merged through the native queue at 2026-10-08 09:59:04 UTC as
dc6dcf61f9ef2f873c6379667108c18a1f983b9f.
Reviewed head: 5c3933e3595e00557e80a3b4d691281189b84439.
All eight required head checks and all eight combined-candidate checks passed.
Actual queue membership was confirmed before inspecting candidate checks.

The maintained merge-integrity reconstruction passed against actual parent
95179adbd0c13c2ca24e25cb160105cc92444544, with complete coverage and no errors:
expected and landed tree both e2e3d9beab675dfb27d23dd1bd162c5b8a140f24.
This also equals the locally tested tree. Local full suite: 3,447 passed.

#1823 closed automatically on merge. The feature branch was lease-deleted at
the reviewed head, its isolated worktree removed without forcing, and its local
branch deleted. After fetch/prune, no local/tracking ref or remote matching ref
remains for codex/add-constricting-ring. The primary main checkout is clean
and equals origin/main at the verified merge; the other 20 worktrees remain.

Final formal and inline review inventories remain empty after the single
successful Copilot request. The cause of that silence is unknown; no independent
approval is claimed. Claude/Shepherd no-review quota receipts and the affected
PR are tracked in #742.
The author adversarial review and repair are documented above.

The two scientific discussions and proposed upstream METPO mapping remain open.
No further skill update is needed for the demonstrated naming failure, which
the current guidance already addresses. This completes this trait's PR cycle;
the broader new-trait discovery goal remains active, 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.

Qualify constricting-ring draft label to distinguish fungal cytokinesis

1 participant