Skip to content

Sequence Detail page surfaces NCBI-record provenance fields

Motivation

A researcher inspecting a BLAST hit (e.g. OZ254605.1) wanted the dashboard's per-accession view to read like the public NCBI Nucleotide record page — i.e. show the sample provenance (isolate, geographic location, collection date, molecule type), the DBLINK cross-references (BioProject / BioSample), the taxonomy lineage, and the publication references. The BLAST results table already deep-links each accession to /sequence/{accession} (SequenceDetail), but that page only rendered a small header (organism / length / molecule / topology / updated) plus a features table — and the features table's Gene / Product column was silently broken.

User-facing change

On the Sequence Detail page (/sequence/{accession}):

  • New "Sample & source" card — pulls the source feature qualifiers NCBI shows: mol_type, isolate, strain, host, geo location (geo_loc_name, falling back to the legacy country qualifier), collection_date, collected_by, isolation_source. Only the qualifiers actually present are rendered.
  • New DBLINK row — BioProject / BioSample (and any other GBSeq_xrefs entries), with outbound links to the NCBI BioProject / BioSample pages.
  • New "Taxonomy" card — the full lineage from GBSeq_taxonomy, rendered as a breadcrumb (Eukaryota › Metazoa › … › Homo).
  • New "References" card — publication title, authors, journal, and a link to PubMed when a PMID is present.
  • Bug fix — the features table Gene / Product column now resolves correctly. The frontend type declared qualifiers as a string map, but the backend returns an ordered [{name, value}] list (to preserve duplicate db_xref entries), so feature.qualifiers.gene was always undefined and every row showed —.

No new page, route, or navigation entry. The data is fetched on demand via the existing GET /api/ncbi/nuccore/{accession}/genbank endpoint when a hit accession is opened, so the NCBI 3 req/s budget is unchanged (one GenBank fetch per opened accession, 24h cached).

API / IaC diff summary

  • Backend api/services/ncbi/nuccore.py: added _parse_xrefs() and an xrefs: list[{dbname, id}] field to the parsed GenBank payload (parsed from GBSeq_xrefs, i.e. the record-level DBLINK block). Additive — no existing field changed.
  • Frontend types web/src/api/ncbi.ts: corrected NuccoreFeature.qualifiers to NuccoreQualifier[], NuccoreGenBank.taxonomy_lineage to string (it is a semicolon-joined string, not an array), NuccoreReference to the real backend shape ({title, journal, authors: string[], pubmed}), and added NuccoreGenBankXref + NuccoreGenBank.xrefs.
  • Frontend web/src/pages/sequence/SequenceDetail.tsx: qualifier-lookup helper, source/DBLINK/taxonomy/references cards, and the Gene/Product column fix.
  • No IaC, Bicep, or Container App changes.

Validation evidence

  • uv run pytest -q api/tests/test_ncbi_nuccore.py → 52 passed (includes the new xrefs assertion in test_fetch_nuccore_genbank_parses_record).
  • uv run ruff check api/services/ncbi/nuccore.py api/tests/test_ncbi_nuccore.py → all checks passed.
  • cd web && npm run build → type-check + production build succeeded (SequenceDetail-*.js emitted, no TS errors).