[KrishnaTO]

Why

Since the canonical ROBOT layout (#154) every rdf:about / rdf:resource in ontologies/ari_t1d.owl is absolute (http://www.aurint.org/ontologies/ari-disease#…, diseases under https://diseases.autoimmuneregistry.org/disease/…). notebook/research-gaps/build_research_gaps.py keyed nodes by about.lstrip("#") and matched #AutoimmuneDisease, so report 12 reported 0 diseases.

What

• build_research_gaps.py resolves every IRI against the file's xml:base (the same urljoin approach as build_uberon_import.py's resolve()) and compares full IRIs (ARI + "AutoimmuneDisease", ARI + "Tissue_Systemic", typed links). It reads the ontology as XML, and nothing else changed. It now finds 228 diseases.
• data/4-reports/12_Research_Gaps.xlsx is regenerated. Compared cell by cell with the committed 2026-10-16 report, every difference comes from curation made since then:
  • +1 disease: ARI:0001168 Primary immune deficiency (Dysregulation). It publishes Disease Type and SNOMED; every other field is Research.
  • ARI:0001091 Evans syndrome: Synonyms went from Research to Published (Update Evans syndrome #156).
  • ARI:0001176 Secondary Raynaud's phenomenon: cited-type items went from 20 to 19 (0 cited either way).
  • The legend and all other rows are unchanged.
• import_columns.py, import_gaps.py are marked HISTORICAL in their docstrings and raise SystemExit, following the pattern import_prevalence.py already uses. They don't only look up nodes: they also splice ari_t1d.owl line by line, anchoring on rdf:about="#…" declarations and writing rdf:resource="#{cls}". Fixing only their lookups would still leave a writer that breaks the canonical layout. Their imports already landed in Import master-list data that report 12 marked 'In repo - import' #141. import_prevalence.py already refuses to run, so it is untouched.
• Changelog entry under ## fix/research-gaps-absolute-iris.

Note on #157

#157 (ontology modules split, still open) renames ari_t1d.owl to ari_disease.owl.

• build_research_gaps.py merges cleanly with Split the ontology into core, disease, symptoms, anatomy and biomarkers modules #157. I ran this branch's builder against Split the ontology into core, disease, symptoms, anatomy and biomarkers modules #157's ari_disease.owl and got the same Summary, Gap Matrix and To Do sheets as the committed report (228 diseases).
• import_columns.py, import_gaps.py and changelog.md conflict, but only textually: Split the ontology into core, disease, symptoms, anatomy and biomarkers modules #157 renames ari_t1d.owl to ari_disease.owl in the docstring paragraph this PR rewrites. Resolve by keeping this PR's paragraph and Split the ontology into core, disease, symptoms, anatomy and biomarkers modules #157's filename.

🤖 Generated with Claude Code