Extend stub-import transform to cover BTO

[realmarcin]

Summary

Promotes BTO to the same SemSQL-backed enriched-stub treatment as NCIT and mesh (introduced by PR #565), so the merged KG carries full label + synonyms + xrefs on both BTO nodes instead of label-only stubs.

Why now

The 2026-05-18 MIM SSSOM republish (PR #564, already merged) added a MIM:Cell_Lysate → BTO:0004304 cell lysate row. Combined with the pre-existing Wound-fluid → BTO:0003114 wound fluid row in mappings/isolation_source_to_ontology.tsv, BTO's in-repo footprint is now 2 IDs — past the threshold where the original ontologies-stubs design (PR #565) opted to leave BTO on the label-only inline path.

Changes

File	Change
kg_microbe/transform_utils/ontologies_stubs/ontologies_stubs_transform.py	Add BTO to STUB_ONTOLOGY_SOURCES (db_filename=bto.db, knowledge_source=infores:bto)
kg_microbe/transform_utils/bacdive/bacdive.py:2991-3007	Extend inline-emit skip-list from {NCIT, mesh} → {NCIT, mesh, BTO} so BacDive defers BTO stub-node emission to the new transform (avoids duplicate node rows). Comment updated.
kg_microbe/utils/isolation_source_mapping_utils.py	STUB_ONTOLOGY_PREFIXES docstring updated to reflect new partitioning (NCIT/mesh/BTO enriched via SemSQL; PRIDE/PCO/GENEPIO/FAO/SNOMED stay on the label-only inline path).
download.yaml	Add bto.db.gz from s3.amazonaws.com/bbop-sqlite (~30 MB)
merge.yaml / merge.no_metatraits.yaml / merge_bakta.yaml	Add data/transformed/ontologies_stubs/bto_nodes.tsv to the ontologies_stubs source filename list in each variant
tests/test_ontologies_stubs.py	Rename test_stub_ontology_sources_covers_ncit_and_mesh → test_stub_ontology_sources_covers_ncit_mesh_bto; assert set is exactly {NCIT, mesh, BTO}

Verification

$ poetry run python -c "from kg_microbe.utils.stub_curie_collection import collect_stub_curies; print(collect_stub_curies(['NCIT','mesh','BTO']))"
{'NCIT': {…73 CURIEs…}, 'mesh': {…95 CURIEs…}, 'BTO': {'BTO:0003114', 'BTO:0004304'}}

$ poetry run pytest tests/test_stub_curie_collection.py tests/test_ontologies_stubs.py -v
13 passed, 1 skipped

$ poetry run python -m ruff check kg_microbe/ tests/
All checks passed!

Test plan

• poetry run kg download — fetches bto.db.gz (~30 MB, one-time; NCIT and mesh DBs already downloaded if PR Selective per-CURIE NCIT/MESH stub-import transform #565 was run end-to-end)
• poetry run kg transform -s ontologies_stubs — should write data/transformed/ontologies_stubs/bto_nodes.tsv with 2 rows + header (BTO:0003114 wound fluid, BTO:0004304 cell lysate), each carrying label/synonyms/xrefs from the BRENDA Tissue Ontology
• poetry run pytest tests/test_ontologies_stubs.py -v — integration test stops being skipped after the transform runs
• After re-merge: spot-check that the two BTO nodes in the merged KG now carry non-empty synonym/xref fields

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR extends the existing ontologies_stubs SemSQL-backed enrichment mechanism (previously for NCIT + MeSH) to also cover BTO, so BTO CURIEs referenced by in-repo mappings resolve to enriched stub nodes (label/synonyms/xrefs) rather than inline label-only stubs.

Changes:

• Add BTO to STUB_ONTOLOGY_SOURCES and skip inline BTO stub-node emission in the BacDive transform to avoid duplicate node rows.
• Add bto.db.gz to download.yaml and include bto_nodes.tsv in the ontologies_stubs source list across merge config variants.
• Update docs/tests to reflect the expanded stub-enrichment scope.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
kg_microbe/transform_utils/ontologies_stubs/ontologies_stubs_transform.py	Adds BTO as a SemSQL-enriched stub ontology source alongside NCIT/MeSH.
kg_microbe/transform_utils/bacdive/bacdive.py	Extends the inline-stub emission skip-list to include BTO (defers to ontologies_stubs).
kg_microbe/utils/isolation_source_mapping_utils.py	Updates documentation describing which stub prefixes are enriched via SemSQL vs emitted inline.
download.yaml	Adds the BTO SemSQL DB (bto.db.gz) to the download manifest.
merge.yaml	Includes bto_nodes.tsv in the ontologies_stubs merge inputs.
merge.no_metatraits.yaml	Includes bto_nodes.tsv in the ontologies_stubs merge inputs (no-metatraits variant).
merge_bakta.yaml	Includes bto_nodes.tsv in the ontologies_stubs merge inputs (bakta variant).
tests/test_ontologies_stubs.py	Updates config test to require {NCIT, mesh, BTO} coverage.

Comments suppressed due to low confidence (1)

kg_microbe/transform_utils/ontologies_stubs/ontologies_stubs_transform.py:88

• The module/class documentation is now out of date with the addition of BTO: the top-level docstring and the __init__ param docs still describe only NCIT/MESH and {ncit,mesh}_nodes.tsv outputs. Please update those docstrings to reflect {ncit,mesh,bto} so future readers don’t miss that BTO is handled by this transform.

class OntologiesStubsTransform(Transform):

    """Emit one labelled stub node per referenced NCIT / MESH / BTO CURIE."""

    def __init__(

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.