PGx: ClinPGx variant-based genes (PGX-008) and --guideline dpwg (PGX-025)

- fetch_sources.sh downloads ClinPGx clinicalVariants (CC BY-SA 4.0); pgx db build adds genotype-only
  variant-based genes (levels 1A-2B, level 3 for the parity list), rsIDs resolved from the manifest's probes:
  70 genes in the DB.
- pgx phenotype --guideline dpwg: CPIC mappings for CPIC genes (as ClinPGx curates DPWG), no phenotype for
  DPWG-only genes (flags dpwg_via_cpic_mapping / dpwg_mapping_not_available). K-029 updated.

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

src/simurg/pipeline/pgx_pheno.py

@@ -17,6 +17,11 @@ from simurg.core.log import Reporter
 from simurg.pgx.phenotype import build_table, gene_phenotype
 
 JSON_SCHEMA = "simurg.pgx/1"
+# PGX-025 DPWG gene coverage. ClinPGx curates DPWG guidelines with the CPIC allele-function and phenotype mappings for
+# genes CPIC covers [DOC: ClinPGx, "Genotype-phenotype mapping for curation of DPWG guidelines"], so those genes use the
+# CPIC tables (flag dpwg_via_cpic_mapping); DPWG-only genes have no public machine-readable mapping here.
+DPWG_GENES = ("CYP1A2", "CYP2B6", "CYP2C19", "CYP2C9", "CYP2D6", "CYP3A4", "CYP3A5", "DPYD", "NUDT15", "SLCO1B1",
+              "TPMT", "UGT1A1", "VKORC1", "F5")  # fmt: skip
 TSV_SCHEMA = "simurg.pgx_tsv/1"
 
 
@@ -65,9 +70,8 @@ def run_phenotype(o: PhenoOptions, rep: Reporter) -> int:
     """``simurg pgx phenotype`` (CLI-034)."""
     from simurg.core.atomic import atomic_write_text
 
-    if o.guideline != "cpic":
-        raise SimurgError("E_PGX_GUIDELINE", f"guideline {o.guideline!r} is not in this PGx DB (CPIC only)",
-                          hint="DPWG tables are not built yet (K-029)")  # fmt: skip
+    if o.guideline not in ("cpic", "dpwg"):
+        raise SimurgError("E_PGX_GUIDELINE", f"unknown guideline {o.guideline!r}", hint="cpic or dpwg")
     src = o.calls / "pgx_calls.json"
     if not src.exists():
         raise SimurgError("E_NO_INPUT", f"{src} not found", hint="run pgx star-allele call first")
@@ -77,9 +81,11 @@ def run_phenotype(o: PhenoOptions, rep: Reporter) -> int:
     meta = db_meta(o.db)
     con = sqlite3.connect(o.db)
     tables = {}
-    for (gene,) in con.execute("SELECT DISTINCT gene FROM phenotypes WHERE guideline = ?", (o.guideline,)):
+    for (gene,) in con.execute("SELECT DISTINCT gene FROM phenotypes WHERE guideline = 'cpic'"):
+        if o.guideline == "dpwg" and gene not in DPWG_GENES:
+            continue
         rows = con.execute("SELECT lookup_key, activity_score, phenotype FROM phenotypes WHERE gene = ? AND "
-                           "guideline = ?", (gene, o.guideline)).fetchall()  # fmt: skip
+                           "guideline = 'cpic'", (gene,)).fetchall()  # fmt: skip
         alle = con.execute("SELECT allele, function, activity FROM alleles WHERE gene = ?", (gene,)).fetchall()
         tables[gene] = build_table(gene, rows, alle)
     tested = {g: [a for (a,) in con.execute("SELECT allele FROM alleles WHERE gene = ? AND defined = 1", (g,))]
@@ -104,6 +110,8 @@ def run_phenotype(o: PhenoOptions, rep: Reporter) -> int:
             elif g.get("call_type") == "copy_number":
                 res, per = None, []
             flags = list(g.get("flags", [])) + (res.flags if res else [])
+            if o.guideline == "dpwg" and g["gene"] in DPWG_GENES:
+                flags.append("dpwg_via_cpic_mapping" if t is not None else "dpwg_mapping_not_available")
             top = g["candidates"][0] if g["candidates"] else {}
             top_alleles = top.get("alleles", [])
             entry = {"gene": g["gene"], "call_type": g.get("call_type", "star_allele"), "genotype": g["genotype"],