PGx: full call detail (PGX-016..020, 027/028, 031), pgx coverage (PGX-009/035), full HGVS

- supporting_variants.tsv, unresolved_variants.tsv, variants_not_covered.tsv with PGX-026 metadata; per-allele
  supporting/unresolved sites for both alleles of every candidate; variant_change vs top, distant_alternative,
  unexplained_variants, alleles_impacted_by_no_call, alleles_tested.
- simurg pgx coverage: callable / not_covered / masked / equivalent_to per allele and manifest.
- HGVS on GRCh38 RefSeq accessions (NC_...:g.).

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01814GKVzZVL3FuF3DeiZXhv

src/simurg/pipeline/pgx_db.py

@@ -113,7 +113,7 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
     """``simurg pgx db build`` (CLI-035)."""
     from simurg.formats.bpm import read_bpm
     from simurg.formats.fasta import ContigMap, Genome
-    from simurg.pgx.db import cpic_definitions, result_basis
+    from simurg.pgx.db import cpic_definitions, full_hgvs, result_basis
 
     cdir = _latest(o.sources, "cpic")
     if cdir is None:
@@ -145,14 +145,14 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
                     continue
                 pos, ref, alt = s.alleles[val]
                 rows["allele_variants"].append(
-                    (gene, d.name, "GRCh38", s.chrom, pos, ref, alt, s.hgvs, s.rsid, lid, val)
+                    (gene, d.name, "GRCh38", s.chrom, pos, ref, alt, full_hgvs(s.chrom, s.hgvs), s.rsid, lid, val)
                 )
             for lid in d.wildcards:
                 rows["allele_wildcards"].append((gene, d.name, lid))
             for pop, f in sorted(d.frequencies.items()):
                 rows["frequencies"].append((gene, d.name, pop, f))
     for lid, s in sorted(sites.items()):
-        rows["sites"].append((s.gene, lid, s.chrom, s.hgvs, s.rsid, s.ref_value))
+        rows["sites"].append((s.gene, lid, s.chrom, full_hgvs(s.chrom, s.hgvs), s.rsid, s.ref_value))
         for val, (pos, ref, alt) in sorted(s.alleles.items()):
             rows["site_values"].append((s.gene, lid, val, pos, ref, alt))
     results = {r["id"]: r for r in t["gene_result"]}
@@ -221,3 +221,71 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
     rep.info(f"PGx DB: {len(defs)} genes, {len(rows['alleles'])} alleles, {len(rows['allele_variants'])} variant "
              f"rows, {len(rows['probe_map'])} probe links → {final} (content sha256 {sha})")  # fmt: skip
     return ExitCode.OK
+
+
+# ------------------------------------------------------------------------------------------------ coverage
+def coverage_rows(db: Path, manifest: str) -> list[dict[str, Any]]:
+    """PGX-009 / PGX-035: per allele, callable | not_covered | masked | equivalent_to:<allele> on a manifest."""
+    import numpy as np
+
+    from simurg.pipeline.pgx_star import load_models
+
+    con = sqlite3.connect(db)
+    links: dict[tuple[str, int, str, str], list[str]] = {}
+    for probe, c, p, r, a in con.execute("SELECT probe_name, chrom, pos, ref, alt FROM probe_map WHERE manifest = ?",
+                                         (manifest,)):  # fmt: skip
+        links.setdefault((c, p, r, a), []).append(probe)
+    masked = {p for (p,) in con.execute("SELECT probe_name FROM masks WHERE manifest = ?", (manifest,))}
+    basis = dict(con.execute("SELECT gene, result_basis FROM genes").fetchall())
+    kinds = {(g, a): (k, d) for g, a, k, d in con.execute("SELECT gene, allele, kind, defined FROM alleles")}
+    con.close()
+    out = []
+    for gene, m in sorted(load_models(db, "global").items()):
+        usable = np.array([bool(links.get(v)) and not all(pr in masked for pr in links[v]) for v in m.variants])
+        masked_v = np.array([bool(links.get(v)) and all(pr in masked for pr in links[v]) for v in m.variants])
+        sig: dict[bytes, list[int]] = {}
+        for h in range(len(m.haplotypes)):
+            sig.setdefault(np.where(usable, m.carries[h], -9).tobytes(), []).append(h)
+        rep = {}
+        for members in sig.values():
+            best = max(members, key=lambda h: (m.freq[h], -h))
+            for h in members:
+                rep[h] = best
+        for h, name in enumerate(m.haplotypes):
+            defining = m.carries[h] == 1
+            probes = sorted({pr for k in np.flatnonzero(defining).tolist() for pr in links.get(m.variants[k], [])})
+            if defining.any() and not (defining & usable).any():
+                status = "masked" if (defining & masked_v).any() else "not_covered"
+            elif rep[h] != h:
+                status = f"equivalent_to:{m.haplotypes[rep[h]]}"
+            else:
+                status = "callable"
+            out.append({"gene": gene, "allele": name, "allele_kind": kinds.get((gene, name), ("", 1))[0],
+                        "result_basis": basis.get(gene, ""), "status": status,
+                        "covering_probes": ";".join(probes)})  # fmt: skip
+    con = sqlite3.connect(db)
+    for gene, allele, kind in con.execute("SELECT gene, allele, kind FROM alleles WHERE defined = 0 ORDER BY 1, 2"):
+        out.append({"gene": gene, "allele": allele, "allele_kind": kind, "result_basis": basis.get(gene, ""),
+                    "status": "not_covered", "covering_probes": ""})  # fmt: skip
+    con.close()
+    return out
+
+
+def run_coverage(bpm: Path, db: Path, out: Path, rep: Reporter) -> int:
+    """``simurg pgx coverage --bpm --db`` (PGX-035)."""
+    from simurg.formats.bpm import read_bpm
+    from simurg.pipeline.pgx_pheno import db_meta, tsv_header, write_meta_tsv
+
+    name = read_bpm(bpm).manifest_name
+    rows = coverage_rows(db, name)
+    if not any(r["covering_probes"] for r in rows):
+        rep.warn("W_PGX_NO_PROBE_MAP", f"the PGx DB has no probe map for {name}; rebuild it with --manifests")
+    out.mkdir(parents=True, exist_ok=True)
+    cols = ["gene", "allele", "allele_kind", "result_basis", "status", "covering_probes"]
+    write_meta_tsv(out / "pgx_coverage.tsv", tsv_header(db, db_meta(db), "none")[:-1], cols, rows)
+    counts: dict[str, int] = {}
+    for r in rows:
+        k = r["status"].split(":")[0]
+        counts[k] = counts.get(k, 0) + 1
+    rep.info(f"{name}: " + ", ".join(f"{k} {v}" for k, v in sorted(counts.items())))
+    return ExitCode.OK