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.

src/simurg/pipeline/pgx_db.py

@@ -254,6 +254,47 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
                 hg = full_hgvs(ps.contig, f"g.{ps.pos}{ps.ref}>{alt}")
                 tag_rows[(gene, allele)] = (gene, allele, "GRCh38", ps.contig, ps.pos, ps.ref, alt, hg, rsid, -1, alt)
                 rows["probe_map"].append((name, ps.name, ps.contig, ps.pos, ps.ref, alt))
+    # ClinPGx clinical-annotation variants of genes no definition source covers: genotype-only variant-based genes
+    cp_src = None
+    cdir2 = _latest(o.sources, "clinpgx")
+    if cdir2 is not None:
+        import zipfile
+
+        from simurg.pgx.db import clinpgx_variant_genes
+
+        cp_src = json.loads((cdir2 / "source.json").read_text())
+        with zipfile.ZipFile(cdir2 / "clinicalVariants.zip") as z:
+            tsv = z.read("clinicalVariants.tsv").decode()
+        known = {g for g, *_ in rows["genes"]} | {g for g, *_ in TAGS}
+        cp_genes = clinpgx_variant_genes(tsv, known)
+        resolved: dict[tuple[str, str], tuple[Any, ...]] = {}
+        want = {rs for v in cp_genes.values() for rs in v}
+        for bpm_path in o.manifests:
+            name = read_bpm(bpm_path).manifest_name
+            for ps in _probe_sites(bpm_path, csvs.get(bpm_path), genome, cmap, {}, rep, want):
+                alts = [a for a in ps.alleles if a != ps.ref]
+                if len(alts) != 1 or ps.ref not in ps.alleles:
+                    continue
+                for rs in re.findall(r"rs\d+", ps.name):
+                    if rs in want:
+                        hg = full_hgvs(ps.contig, f"g.{ps.pos}{ps.ref}>{alts[0]}")
+                        resolved.setdefault((rs, ps.contig), (ps.contig, ps.pos, ps.ref, alts[0], hg))
+                        rows["probe_map"].append((name, ps.name, ps.contig, ps.pos, ps.ref, alts[0]))
+        by_rs = {rs: v for (rs, _), v in resolved.items()}
+        lid2 = 950_000_000
+        for gene in sorted(cp_genes):
+            chrom = None
+            rows["alleles"].append((gene, "Reference", "reference", None, None, 0, 1))
+            for rs in sorted(cp_genes[gene], key=lambda r: int(r[2:])):
+                v = by_rs.get(rs)
+                rows["alleles"].append((gene, rs, "rsid_variant_allele", None, None, 0, int(v is not None)))
+                if v is not None:
+                    lid2 += 1
+                    chrom = v[0]
+                    rows["allele_variants"].append((gene, rs, "GRCh38", v[0], v[1], v[2], v[3], v[4], rs, lid2, v[3]))
+            rows["genes"].append((gene, "variant_based", chrom))
+        rep.info(f"ClinPGx: {len(cp_genes)} variant-based gene(s), {len(by_rs)} of {len(want)} rsID(s) resolved "
+                 f"on the manifest(s)")  # fmt: skip
     if tag_rows:  # PGX-007: tag alleles become defined single-variant alleles
         rows["allele_variants"] += list(tag_rows.values())
         rows["alleles"] = [(g, a, k, f, act, sv, 1 if (g, a) in tag_rows else d) for g, a, k, f, act, sv, d in
@@ -270,7 +311,10 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
                                            "license": csrc.get("license")}]
                                          + ([{"name": "PharmVar", "version": pv_src["version"],
                                               "retrieved": pv_src["retrieved"], "license": pv_src.get("license")}]
-                                            if pv_src else [])),
+                                            if pv_src else [])
+                                         + ([{"name": "ClinPGx clinical variants", "version": cp_src["version"],
+                                              "retrieved": cp_src["retrieved"], "license": cp_src.get("license")}]
+                                            if cp_src else [])),
             "phenotype_sources": json.dumps([{"name": "CPIC", "version": csrc["version"]}]),
             "cpic_files_sha256": json.dumps(csrc.get("files", {}), sort_keys=True),
             "manifests": json.dumps(sorted(manifests)), "definition_notes": json.dumps(problems)}  # fmt: skip
@@ -301,8 +345,9 @@ def run_db_build(o: PgxDbOptions, rep: Reporter) -> int:
         rep.info(f"identical PGx DB already present: {final}")
         return ExitCode.OK
     tmp.replace(final)
-    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
+    n_g, n_a, n_v, n_p = (len(rows[k]) for k in ("genes", "alleles", "allele_variants", "probe_map"))
+    rep.info(f"PGx DB: {n_g} genes, {n_a} alleles, {n_v} variant rows, {n_p} probe links → {final} "
+             f"(content sha256 {sha})")  # fmt: skip
     return ExitCode.OK