2026-03-23: 3.0.5:
- adjust aggregators to handle individual IDC cases mapping to multiple distinct GC participants
- harden GC extractor to handle new array values coming in encoded as strings (race, ethnicity)
- add support for GC consent_group, investigator entities
- harden PDC extractor to handle new ambiguities in project associations introduced by new CPTAC3 reprocessing data of subjects originating elsewhere
- add 'concept' to column_metadata to flag harmonized columns
- harden OBO processor
- add postgres tsvector data and keyword associations to support new global search
- profile all keyword frequencies per release
- fill in all missing file, subject aliases, partnered with nulls, into file_describes_subject to strengthen API assumptions designed to improve query efficiency
- process controlled term sets defined by automatically-loaded external ontologies and by custom CDA maps, including processing of related terms like synonyms, slim-map category assignments and containing/superset (hierarchical) concepts
- add referential indirection for all controlled terms for better master data management and leaner data tables
- adjust PDC extractor delays to minimize run-breakage risk
- expand slim-map format and add count profiling for all categories during each release
- add anatomic_site slim