VariantEffect / mavedb-api / 31848609802

89%

main: 89%

Build Type

push

github

Committed by bencap

Commit Message

feat(scripts): cluster score-set pipeline cohorts by gene symbol

Replace taxonomy/organism cohort filtering with gene-symbol clustering
so run_score_set_pipelines fills concurrency slots one gene at a time,
maximizing ClinGen Allele Registry cache reuse within its 24h TTL.

- Add cluster_cohort/CohortEntry, replacing normalize_gene/order_cohort/
  build_cohort_items; symbols come from mapped_hgnc_name, falling back
  to the first word of the target gene name
- plan_enqueue now spends slots in cluster order so a window stays on
  one gene until it's exhausted, instead of splitting across genes
- Add --gene to filter the cohort by symbol, and --emit-cohorts/
  --cohort-out to preview clusters and write per-cluster URN files
  consumable by --urns-file
- Drop --taxonomy-id/--organism and their TargetSequence/Taxonomy join,
  which was silently excluding accession-based score sets from the
  cohort
- Replace classify_status's terminal/in_flight split with a single
  _IN_FLIGHT_STATUSES set plus an exhaustiveness test
- Update make_score_set fixture with mapped_hgnc_names/num_variants
  and rewrite affected tests

Coverage Stats

69 of 121 new or added lines in 1 file covered. (57.02%)

1 existing line in 1 file now uncovered.

13833 of 15553 relevant lines covered (88.94%)

0.89 hits per line