gnomancer 0.3.2__tar.gz → 0.4.0__tar.gz

gnomancer-0.4.0/PKG-INFO

@@ -0,0 +1,92 @@
+Metadata-Version: 2.4
+Name: gnomancer
+Version: 0.4.0
+Summary: Local, private deep dive into a consumer raw DNA file (AncestryDNA, 23andMe, MyHeritage, FTDNA, LivingDNA, VCF): QC, haplogroups, traits, pharmacogenomics, one HTML report.
+License: MIT
+Project-URL: Homepage, https://github.com/codebase/gnomancer
+Project-URL: Issues, https://github.com/codebase/gnomancer/issues
+Keywords: dna,genetics,23andme,ancestrydna,haplogroup,pharmacogenomics,clinvar,mcp
+Classifier: Programming Language :: Python :: 3
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: snps>=2.9
+Requires-Dist: pandas>=2.0
+Requires-Dist: numpy>=1.24
+Provides-Extra: dev
+Requires-Dist: pytest>=8; extra == "dev"
+Requires-Dist: mcp>=2; extra == "dev"
+Provides-Extra: mcp
+Requires-Dist: mcp>=2; extra == "mcp"
+Dynamic: license-file
+
+<h1 align="center">
+  <img src="docs/logo.png" width="160" alt=""><br>
+  gnomancer
+</h1>
+
+<p align="center"><strong>Your DNA, decoded on your own machine.</strong><br>
+<a href="https://gnomancer.com">gnomancer.com</a> ·
+<a href="https://gnomancer.com/get-your-data/">get your raw data</a> ·
+<a href="https://gnomancer.com/install/">install</a> ·
+<a href="https://pypi.org/project/gnomancer/">PyPI</a> ·
+<a href="https://registry.modelcontextprotocol.io">MCP Registry</a> ·
+<a href="https://discord.gg/aGjqzqErQH">Discord</a></p>
+
+<p align="center">
+<a href="https://pypi.org/project/gnomancer/"><img src="https://img.shields.io/pypi/v/gnomancer?color=7c3aed" alt="PyPI"></a>
+<a href="https://github.com/codebase/gnomancer/actions/workflows/ci.yml"><img src="https://github.com/codebase/gnomancer/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
+<a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-7c3aed" alt="MIT"></a>
+</p>
+
+Take the raw file you already downloaded from **AncestryDNA, 23andMe, MyHeritage, FamilyTreeDNA or LivingDNA** and turn it into one readable report. Everything runs on your computer. Nothing is uploaded.
+
+## Quick start
+
+```bash
+pipx install gnomancer      # Python 3.10+; Java 11+ for the haplogroup and PharmCAT steps
+gnomancer setup             # fetches the open-source tools and reference files once (~250 MB)
+gnomancer ~/Downloads/AncestryDNA.zip
+```
+
+Open `AncestryDNA_analysis/REPORT.html`. No Python? `curl -fsSL https://gnomancer.com/install.sh | bash` does it all. Windows: [via WSL](https://gnomancer.com/install/).
+
+![demo report](docs/demo-report.png)
+
+## What you get
+
+- **Plain language first.** An at-a-glance digest, then a narrative your own AI assistant can write from a prompt the run gives you.
+- **Lineages.** Y haplogroup (yhaplo, plus Yleaf on current YFull/FTDNA/ISOGG trees) and mtDNA haplogroup (Haplogrep3), each with population context.
+- **Traits and ancestry markers.** About 110 well-replicated sites: pigmentation, lactose, caffeine, bitter taste, blood group and more.
+- **Health, carefully.** Carrier sites, common risk variants, a ClinVar scan of every typed site with plausibility scoring, and PharmCAT pharmacogenomics. All collapsed by default so nobody sees a result they didn't ask for.
+- **Stays current.** `gnomancer watch` re-checks ClinVar each month and tells you only what was reclassified at sites you carry.
+- **Honest limits.** Every section says what the chip can and cannot tell you.
+
+## Use it from your AI assistant
+
+gnomancer is also an MCP server. Connect it and ask "what does my report say about caffeine?"
+
+```bash
+claude mcp add gnomancer -- uvx gnomancer serve          # Claude Code; Claude Desktop/Cursor: command "uvx", args ["gnomancer","serve"]
+```
+
+Only derived results reach the assistant, never the raw file. Health tables are separate tools fetched only on request. The repo also works as a Claude plugin (`skills/gnomancer/SKILL.md`).
+
+## Privacy
+
+Your genotype file never leaves your machine. The only network traffic is the one-time download of public tools and references. The optional `--ai` flag sends the structured results (not the file) to the Anthropic API to write the narrative, and only when you ask.
+
+## Not medical advice
+
+A genotyping array reads about 0.02% of the genome at pre-chosen sites, and its calls at rare sites are wrong often enough that a clinical lab found [40% did not confirm](https://doi.org/10.1038/gim.2018.38). gnomancer tells you what the chip says and how far to trust it. Confirm anything that matters with a clinical-grade test and a clinician or genetic counselor.
+
+## More
+
+[Discord community](https://discord.gg/aGjqzqErQH) · [CLI options](docs/cli.md) · [vendors and quirks](docs/vendors.md) · [the marker panel](docs/marker-panel.md) · [tools under the hood](docs/tools.md) · [development](docs/development.md) · [MCP registry](docs/mcp-registry.md) · [contributing](CONTRIBUTING.md)
+
+MIT licensed. The bundled tools keep their own licenses.
+
+<sub>mcp-name: com.gnomancer/gnomancer</sub>

gnomancer-0.4.0/README.md

@@ -0,0 +1,67 @@
+<h1 align="center">
+  <img src="docs/logo.png" width="160" alt=""><br>
+  gnomancer
+</h1>
+
+<p align="center"><strong>Your DNA, decoded on your own machine.</strong><br>
+<a href="https://gnomancer.com">gnomancer.com</a> ·
+<a href="https://gnomancer.com/get-your-data/">get your raw data</a> ·
+<a href="https://gnomancer.com/install/">install</a> ·
+<a href="https://pypi.org/project/gnomancer/">PyPI</a> ·
+<a href="https://registry.modelcontextprotocol.io">MCP Registry</a> ·
+<a href="https://discord.gg/aGjqzqErQH">Discord</a></p>
+
+<p align="center">
+<a href="https://pypi.org/project/gnomancer/"><img src="https://img.shields.io/pypi/v/gnomancer?color=7c3aed" alt="PyPI"></a>
+<a href="https://github.com/codebase/gnomancer/actions/workflows/ci.yml"><img src="https://github.com/codebase/gnomancer/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
+<a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-7c3aed" alt="MIT"></a>
+</p>
+
+Take the raw file you already downloaded from **AncestryDNA, 23andMe, MyHeritage, FamilyTreeDNA or LivingDNA** and turn it into one readable report. Everything runs on your computer. Nothing is uploaded.
+
+## Quick start
+
+```bash
+pipx install gnomancer      # Python 3.10+; Java 11+ for the haplogroup and PharmCAT steps
+gnomancer setup             # fetches the open-source tools and reference files once (~250 MB)
+gnomancer ~/Downloads/AncestryDNA.zip
+```
+
+Open `AncestryDNA_analysis/REPORT.html`. No Python? `curl -fsSL https://gnomancer.com/install.sh | bash` does it all. Windows: [via WSL](https://gnomancer.com/install/).
+
+![demo report](docs/demo-report.png)
+
+## What you get
+
+- **Plain language first.** An at-a-glance digest, then a narrative your own AI assistant can write from a prompt the run gives you.
+- **Lineages.** Y haplogroup (yhaplo, plus Yleaf on current YFull/FTDNA/ISOGG trees) and mtDNA haplogroup (Haplogrep3), each with population context.
+- **Traits and ancestry markers.** About 110 well-replicated sites: pigmentation, lactose, caffeine, bitter taste, blood group and more.
+- **Health, carefully.** Carrier sites, common risk variants, a ClinVar scan of every typed site with plausibility scoring, and PharmCAT pharmacogenomics. All collapsed by default so nobody sees a result they didn't ask for.
+- **Stays current.** `gnomancer watch` re-checks ClinVar each month and tells you only what was reclassified at sites you carry.
+- **Honest limits.** Every section says what the chip can and cannot tell you.
+
+## Use it from your AI assistant
+
+gnomancer is also an MCP server. Connect it and ask "what does my report say about caffeine?"
+
+```bash
+claude mcp add gnomancer -- uvx gnomancer serve          # Claude Code; Claude Desktop/Cursor: command "uvx", args ["gnomancer","serve"]
+```
+
+Only derived results reach the assistant, never the raw file. Health tables are separate tools fetched only on request. The repo also works as a Claude plugin (`skills/gnomancer/SKILL.md`).
+
+## Privacy
+
+Your genotype file never leaves your machine. The only network traffic is the one-time download of public tools and references. The optional `--ai` flag sends the structured results (not the file) to the Anthropic API to write the narrative, and only when you ask.
+
+## Not medical advice
+
+A genotyping array reads about 0.02% of the genome at pre-chosen sites, and its calls at rare sites are wrong often enough that a clinical lab found [40% did not confirm](https://doi.org/10.1038/gim.2018.38). gnomancer tells you what the chip says and how far to trust it. Confirm anything that matters with a clinical-grade test and a clinician or genetic counselor.
+
+## More
+
+[Discord community](https://discord.gg/aGjqzqErQH) · [CLI options](docs/cli.md) · [vendors and quirks](docs/vendors.md) · [the marker panel](docs/marker-panel.md) · [tools under the hood](docs/tools.md) · [development](docs/development.md) · [MCP registry](docs/mcp-registry.md) · [contributing](CONTRIBUTING.md)
+
+MIT licensed. The bundled tools keep their own licenses.
+
+<sub>mcp-name: com.gnomancer/gnomancer</sub>

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/__init__.py

@@ -1,2 +1,2 @@
 """gnomancer: vendor-agnostic analysis of consumer raw DNA files."""
-__version__ = "0.3.2"
+__version__ = "0.4.0"

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/cli.py

@@ -48,7 +48,10 @@ def main(argv=None):
     if argv[:1] == ["setup"]:
         from .setup_cmd import run as setup
         return setup(argv[1:])
-    ap = argparse.ArgumentParser(prog="gnomancer", description="Analyze a consumer raw DNA file locally. Subcommands: `gnomancer setup` (download tools), `gnomancer serve` (MCP server).")
+    if argv[:1] == ["watch"]:
+        from .watch import run as watch
+        return watch(argv[1:])
+    ap = argparse.ArgumentParser(prog="gnomancer", description="Analyze a consumer raw DNA file locally. Subcommands: `gnomancer setup` (download tools), `gnomancer serve` (MCP server), `gnomancer watch REPORT_DIR` (ClinVar reclassification check).")
     ap.add_argument("input", help="raw data file (.txt/.csv/.tsv/.vcf, optionally .zip/.gz)")
     ap.add_argument("--out", default=None, help="output directory (default: <input>_analysis)")
     ap.add_argument("--markers", default=str(DATA / "markers.tsv"), help="marker panel TSV")
@@ -108,6 +111,9 @@ def main(argv=None):
         cv = clinvar_mod.scan(df)
         if cv["available"] and cv["hits"] is not None and not cv["hits"].empty:
             cv["hits"].to_csv(out / "clinvar-hits.tsv", sep="\t", index=False)
+        if cv["available"]:
+            from .watch import write_snapshot
+            write_snapshot(out, cv, str(inp.resolve()))
     cv_text = clinvar_mod.interpret(cv) if not a.no_clinvar else cv["note"]
     pc = dict(available=False, calls=None, error="skipped (--no-pharmcat)", stats={})
     if not a.no_pharmcat:
@@ -134,7 +140,8 @@ def main(argv=None):
             (out / "summary.md").write_text(narrative + "\n")
         except Exception as e:  # noqa: BLE001
             loaded.notes.append(f"AI summary failed: {e}")
-    extra = dict(clinvar=cv, clinvar_text=cv_text, pharmcat=pc, pharmcat_text=pc_text, glance=glance, narrative=narrative)
+    pop = (ancestry or {}).get("superpopulation_code") or "EUR"
+    extra = dict(clinvar=cv, clinvar_text=cv_text, pharmcat=pc, pharmcat_text=pc_text, glance=glance, narrative=narrative, pop=pop)
     rep = write(out / "REPORT.md", loaded, qc, ct, roh, roh_text, panel, combo, y, mt, ancestry, extra)
     html = write_html(out / "REPORT.html", loaded, qc, ct, roh, roh_text, panel, combo, y, mt, ancestry, input_name=inp.name, extra=extra)
     json.dump(dict(input=str(inp.resolve()), args=[str(Path(x).resolve()) if Path(x).exists() else x for x in run_args],

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/clinvar.py

@@ -45,7 +45,8 @@ def _info(s: str) -> dict:
 
 def scan(df: pd.DataFrame, vcf: Path = DEFAULT_VCF) -> dict:
     """df: normalized SNP table (chrom, pos, genotype). Returns dict with hits table and coverage stats."""
-    res = dict(available=vcf.exists(), vcf=str(vcf), sites_in_clinvar=0, pathogenic_sites_typed=0, hits=pd.DataFrame(), note="")
+    res = dict(available=vcf.exists(), vcf=str(vcf), sites_in_clinvar=0, pathogenic_sites_typed=0, hits=pd.DataFrame(), note="",
+               carried={}, release=None)
     if not vcf.exists():
         res["note"] = "ClinVar VCF not found; run `gnomancer setup` (downloads ~200 MB from NCBI)."
         return res
@@ -54,10 +55,13 @@ def scan(df: pd.DataFrame, vcf: Path = DEFAULT_VCF) -> dict:
     for chrom, pos, gt in zip(called.chrom, called.pos, called.genotype):
         key[(chrom, int(pos))] = gt
     hits = []
+    carried_sites = {}
     n_in = n_path = 0
     with gzip.open(vcf, "rt") as f:
         for line in f:
             if line[0] == "#":
+                if line.startswith("##fileDate="):
+                    res["release"] = line.strip().split("=", 1)[1]
                 continue
             p = line.rstrip("\n").split("\t", 8)
             chrom = p[0]
@@ -70,6 +74,10 @@ def scan(df: pd.DataFrame, vcf: Path = DEFAULT_VCF) -> dict:
             n_in += 1
             info = _info(p[7])
             sig = info.get("CLNSIG", "")
+            ref_, alts_ = p[3], p[4].split(",")
+            if len(ref_) == 1 and all(len(a) == 1 for a in alts_) and any(a in alts_ for a in gt) and all(a in alts_ or a == ref_ for a in gt):
+                carried_sites[f"{chrom}:{p[1]}:{p[4]}"] = dict(rsid=("rs" + info["RS"]) if info.get("RS") else p[2], gene=info.get("GENEINFO", "").split(":")[0],
+                                                       sig=sig, rev=info.get("CLNREVSTAT", ""), gt=gt)
             if not any(s in sig for s in PATHOGENIC) or sig.startswith("Benign") or sig.startswith("Likely_benign"):
                 continue
             n_path += 1
@@ -109,6 +117,7 @@ def scan(df: pd.DataFrame, vcf: Path = DEFAULT_VCF) -> dict:
                              condition=_conditions(info.get("CLNDN", "")),
                              variant=info.get("CLNHGVS", ""), alt_af=af, confidence=confidence, flags="; ".join(flags)))
     res["sites_in_clinvar"], res["pathogenic_sites_typed"] = n_in, n_path
+    res["carried"] = carried_sites
     out = pd.DataFrame(hits)
     if not out.empty:
         out["conf_rank"] = out.confidence.map({"high": 0, "medium": 1, "low": 2})

gnomancer-0.4.0/gnomancer/data/marker_frequencies.tsv

@@ -0,0 +1,113 @@
+# Effect-allele frequency per 1000 Genomes phase 3 super-population (Ensembl GRCh37 REST). Regenerate with examples/fetch_frequencies.py
+rsid	AFR	AMR	EAS	EUR	SAS	note
+rs1426654	0.0741	0.5893	0.0119	0.9970	0.6851	direct
+rs16891982	0.0356	0.4640	0.0060	0.9384	0.0593	direct
+rs3827760	0.0030	0.3919	0.8730	0.0109	0.0133	direct
+rs2814778	0.9637	0.0778	0.0000	0.0060	0.0000	direct
+rs4988235	0.0272	0.2161	0.0000	0.5080	0.1135	direct
+rs182549	0.0272	0.2205	0.0000	0.5089	0.1196	direct
+rs671	0.0015	0.0029	0.1736	0.0000	0.0000	direct
+rs1229984	0.0015	0.0576	0.6974	0.0288	0.0204	direct
+rs12913832	0.0280	0.2017	0.0020	0.6362	0.0706	direct
+rs1800407	0.0015	0.0317	0.0010	0.0755	0.0266	direct
+rs12896399	0.0212	0.2522	0.3472	0.4304	0.3272	direct
+rs1042602	0.0121	0.2378	0.0010	0.3718	0.0634	direct
+rs1393350	0.0083	0.1225	0.0030	0.2435	0.0542	direct
+rs12203592	0.0083	0.0720	0.0000	0.1163	0.0061	direct
+rs1805007	0.0030	0.0159	0.0010	0.0716	0.0051	direct
+rs1805008	0.0038	0.0029	0.0000	0.0616	0.0041	direct
+rs1805009	0.0015	0.0101	0.0000	0.0080	0.0000	direct
+rs1805006	0.0008	0.0014	0.0000	0.0099	0.0010	direct
+rs11547464	0.0008	0.0058	0.0000	0.0089	0.0000	direct
+rs1110400	0.0000	0.0086	0.0000	0.0080	0.0020	direct
+rs885479	0.0068	0.3156	0.6161	0.0696	0.0399	direct
+rs11803731	0.9955	0.9063	0.9990	0.7714	0.9366	direct
+rs17822931	0.0121	0.1398	0.7798	0.1362	0.4816	direct
+rs4481887	0.0998	0.2176	0.1379	0.2773	0.1022	direct
+rs10427255	0.4130	0.3732	0.4415	0.4950	0.4489	direct
+rs1815739	0.1150	0.5720	0.4435	0.4344	0.5869	direct
+rs762551	0.4380	0.2421	0.3274	0.3201	0.4652	direct
+rs2472297	0.0144	0.0893	0.0000	0.2147	0.0327	direct
+rs4410790	0.4402	0.3732	0.4137	0.6153	0.4724	direct
+rs713598						no 1000G data
+rs1726866						no 1000G data
+rs10246939						no 1000G data
+rs72921001	0.7927	0.6268	0.6865	0.6431	0.5746	direct
+rs9939609	0.4939	0.2622	0.1687	0.4135	0.2883	direct
+rs17782313	0.2776	0.1340	0.1865	0.2396	0.3200	direct
+rs16969968	0.0234	0.2089	0.0268	0.3658	0.1820	direct
+rs4680	0.2806	0.3775	0.2798	0.5000	0.4407	direct
+rs6265	0.0106	0.1527	0.4881	0.1968	0.2025	direct
+rs1799971	0.0091	0.2003	0.3929	0.1620	0.4182	direct
+rs601338	0.4909	0.3415	0.0040	0.4414	0.2832	direct
+rs2180439	0.4463	0.5317	0.4067	0.4175	0.5307	direct
+rs53576	0.1936	0.3559	0.6498	0.3509	0.4489	direct
+rs1800497	0.3850	0.3112	0.4058	0.1879	0.3149	direct
+rs8176719	0.7057	0.7680	0.6200	0.6054	0.5992	indel
+rs8176746						no 1000G data
+rs1050828						no 1000G data
+rs429358	0.2678	0.1037	0.0863	0.1551	0.0869	direct
+rs7412	0.1029	0.0476	0.1002	0.0626	0.0440	direct
+rs6025	0.0000	0.0101	0.0000	0.0119	0.0112	direct
+rs1799963	0.0000	0.0144	0.0000	0.0080	0.0000	direct
+rs1800562	0.0023	0.0216	0.0000	0.0427	0.0020	direct
+rs1799945	0.0113	0.1153	0.0288	0.1720	0.0706	direct
+rs1800730	0.0000	0.0043	0.0000	0.0159	0.0010	direct
+rs1801133	0.0900	0.4741	0.2956	0.3648	0.1186	direct
+rs1801131	0.1513	0.1513	0.2192	0.3131	0.4172	direct
+rs2187668	0.0537	0.1182	0.0516	0.1064	0.0787	direct
+rs7454108						no 1000G data
+rs3135388	0.0272	0.0504	0.0060	0.1203	0.0174	direct
+rs2476601	0.0030	0.0360	0.0000	0.0944	0.0133	direct
+rs2066844	0.0023	0.0245	0.0000	0.0507	0.0010	direct
+rs2066845	0.0000	0.0130	0.0000	0.0099	0.0041	direct
+rs7903146	0.2602	0.2349	0.0228	0.3171	0.2986	direct
+rs1061170	0.3623	0.2334	0.0486	0.3618	0.2873	direct
+rs10490924	0.2458	0.2478	0.4038	0.1948	0.3425	direct
+rs1333049	0.2133	0.4553	0.5367	0.4722	0.4908	direct
+rs10455872	0.0023	0.0331	0.0050	0.0736	0.0061	direct
+rs3798220	0.0045	0.2161	0.0883	0.0099	0.0020	direct
+rs11136000	0.5741	0.3573	0.2192	0.3877	0.2822	direct
+rs75932628	0.0000	0.0043	0.0000	0.0050	0.0020	direct
+rs76763715						no 1000G data
+rs34637584	0.0000	0.0014	0.0000	0.0000	0.0000	direct
+rs2802292	0.8290	0.3905	0.3056	0.4334	0.5613	direct
+rs1800629	0.1195	0.0692	0.0585	0.1342	0.0532	direct
+rs113993960						indel alleles not comparable
+rs333						indel alleles not comparable
+rs80357906						no 1000G data
+rs80359550						no 1000G data
+rs11571833	0.0008	0.0029	0.0000	0.0109	0.0082	direct
+rs334	0.0998	0.0072	0.0000	0.0000	0.0000	direct
+rs33930165	0.0129	0.0000	0.0000	0.0000	0.0000	direct
+rs33950507	0.0000	0.0000	0.0079	0.0000	0.0061	direct
+rs1799853	0.0083	0.0994	0.0010	0.1243	0.0348	direct
+rs1057910	0.0023	0.0375	0.0337	0.0726	0.1094	direct
+rs28371686	0.0166	0.0014	0.0000	0.0000	0.0000	direct
+rs9332131						indel alleles not comparable
+rs28371685	0.0242	0.0014	0.0000	0.0020	0.0010	direct
+rs4244285	0.1702	0.1052	0.3125	0.1451	0.3579	direct
+rs4986893	0.0023	0.0000	0.0556	0.0000	0.0123	direct
+rs12248560	0.2352	0.1196	0.0149	0.2237	0.1360	direct
+rs9923231	0.0545	0.4107	0.8849	0.3877	0.1452	direct
+rs9934438	0.0545	0.4121	0.8849	0.3877	0.1452	direct
+rs2108622	0.0825	0.2378	0.2143	0.2903	0.4131	direct
+rs3892097						no 1000G data
+rs1065852						no 1000G data
+rs28371725						no 1000G data
+rs776746	0.8200	0.2032	0.2867	0.0567	0.3323	direct
+rs3745274	0.3744	0.3732	0.2153	0.2356	0.3814	direct
+rs4149056	0.0136	0.1340	0.1230	0.1610	0.0429	direct
+rs2231142	0.0129	0.1412	0.2907	0.0944	0.0971	direct
+rs3918290	0.0008	0.0014	0.0000	0.0050	0.0082	direct
+rs67376798	0.0008	0.0029	0.0000	0.0070	0.0010	direct
+rs887829	0.4932	0.3790	0.1300	0.2982	0.4366	direct
+rs4148323	0.0008	0.0115	0.1379	0.0070	0.0174	direct
+rs1801280	0.2920	0.3617	0.0377	0.4493	0.3466	direct
+rs1799930	0.2375	0.1715	0.2560	0.2823	0.3599	direct
+rs1799931	0.0287	0.1124	0.1796	0.0229	0.0685	direct
+rs1495741	0.4191	0.3501	0.5159	0.2435	0.2209	direct
+rs1800460	0.0030	0.0403	0.0000	0.0278	0.0041	direct
+rs1142345	0.0666	0.0576	0.0218	0.0288	0.0174	direct
+rs2395029	0.0045	0.0202	0.0119	0.0437	0.0930	direct
+rs1800566	0.1755	0.3329	0.4187	0.2107	0.3579	direct

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/mcp_server.py

@@ -33,7 +33,7 @@ Rules: the file never leaves the machine and you must not paste genotype rows an
 rare-site calls need clinical confirmation. Never diagnose. If the user has no file yet, read the
 gnomancer://guide/get-your-data resource."""
 
-server = MCPServer("gnomancer", version="0.3.2", instructions=INSTRUCTIONS, website_url="https://gnomancer.com")
+server = MCPServer("gnomancer", version="0.4.0", instructions=INSTRUCTIONS, website_url="https://gnomancer.com")
 _state: dict = {"report_dir": os.environ.get("GNOMANCER_REPORT"), "df_cache": {}}
 
 
@@ -150,6 +150,19 @@ def pharmacogenomics(report_dir: str | None = None) -> dict:
     return s.get("pharmcat") or {}
 
 
+@server.tool()
+def clinvar_watch(report_dir: str | None = None, force: bool = False) -> str:
+    """Re-check ClinVar for this report: downloads a new ClinVar release if NCBI published one, rescans the user's
+    sites and reports only classification changes (newly pathogenic, downgraded, review-status changes). The first
+    run writes a baseline. Health-sensitive: run when the user asks."""
+    d = _report_dir(report_dir)
+    args = ["watch", str(d)] + (["--force"] if force else [])
+    p = _cli(args)
+    if p.returncode != 0:
+        raise RuntimeError((p.stderr or p.stdout)[-1200:])
+    return (p.stdout or "").strip() or (p.stderr or "").strip()
+
+
 @server.tool()
 def narrative_prompt(report_dir: str | None = None) -> str:
     """The prompt for writing the plain-language narrative that goes at the top of the report. Follow it exactly,

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/panel.py

@@ -17,6 +17,30 @@ import pandas as pd
 
 COMP = {"A": "T", "T": "A", "C": "G", "G": "C", "I": "I", "D": "D"}
 CATEGORY_ORDER = ["ancestry", "trait", "health", "carrier", "pgx"]
+POPS = ("AFR", "AMR", "EAS", "EUR", "SAS")
+POP_LABEL = dict(AFR="African", AMR="Admixed American", EAS="East Asian", EUR="European", SAS="South Asian")
+FREQ_FILE = Path(__file__).resolve().parent / "data" / "marker_frequencies.tsv"
+
+
+def load_frequencies(path: Path = FREQ_FILE) -> dict:
+    """rsid -> {pop: effect-allele frequency} from the bundled 1000 Genomes table."""
+    out = {}
+    if not path.exists():
+        return out
+    with open(path, newline="") as f:
+        lines = [l for l in f if l.strip() and not l.startswith("#")]
+    for r in csv.DictReader(lines, delimiter="\t"):
+        d = {p: float(r[p]) for p in POPS if r.get(p)}
+        if d:
+            out[r["rsid"]] = d
+    return out
+
+
+def genotype_frequency(effect_count: int, ploidy: int, p: float) -> float:
+    """Share of a population expected to carry this genotype (Hardy-Weinberg)."""
+    if ploidy == 1:
+        return p if effect_count == 1 else 1 - p
+    return {0: (1 - p) ** 2, 1: 2 * p * (1 - p), 2: p * p}[effect_count]
 
 
 def load_markers(path: Path) -> list[dict]:
@@ -37,13 +61,14 @@ def _lookup(df: pd.DataFrame, rsid: str):
     return None
 
 
-def evaluate(df: pd.DataFrame, markers: list[dict]) -> pd.DataFrame:
+def evaluate(df: pd.DataFrame, markers: list[dict], frequencies: dict | None = None) -> pd.DataFrame:
+    frequencies = load_frequencies() if frequencies is None else frequencies
     out = []
     for m in markers:
         row = _lookup(df, m["rsid"])
         rec = dict(rsid=m["rsid"], category=m["category"], gene=m["gene"], label=m["label"],
                    ref=m["ref"], effect=m["effect"], chrom="", pos="", genotype="", effect_count="",
-                   interpretation="", note=m.get("note", ""))
+                   interpretation="", note=m.get("note", ""), **{f"gf_{p}": None for p in POPS})
         if row is None:
             rec["genotype"] = "not on array"
             out.append(rec)
@@ -71,6 +96,8 @@ def evaluate(df: pd.DataFrame, markers: list[dict]) -> pd.DataFrame:
         n = g.count(m["effect"])
         ploidy = len(g)
         rec["effect_count"] = f"{n}/{ploidy}"
+        for pop, af in (frequencies.get(m["rsid"]) or {}).items():
+            rec[f"gf_{pop}"] = round(genotype_frequency(n, ploidy, af), 4)
         key = {0: "interp0", 1: "interp1", 2: "interp2"}[n if ploidy == 2 else (2 if n == 1 else 0)]
         text = m.get(key) or f"{n} copy(ies) of effect allele {m['effect']}"
         if ploidy == 1:

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/report.py

@@ -5,7 +5,7 @@ from pathlib import Path
 
 import pandas as pd
 
-from .panel import CATEGORY_ORDER
+from .panel import CATEGORY_ORDER, POPS, POP_LABEL
 
 DISCLAIMER = (
     "Computed locally from a consumer genotyping array. The array reports a fixed list of common sites, "
@@ -104,7 +104,10 @@ def write(out: Path, loaded, qc: dict, chr_table: pd.DataFrame, roh: pd.DataFram
     for name, val, note in combos:
         L.append(f"- **{name}:** {val}" + (f" — {note}" if note else ""))
     L.append("\n## 8. Marker panel\n")
-    L.append("Effect-allele counts are on the GRCh37 forward strand. `not on array` means the vendor's chip does not type that site.\n")
+    pop = extra.get("pop") or "EUR"
+    L.append(f"Effect-allele counts are on the GRCh37 forward strand. `not on array` means the vendor's chip does not type that site. "
+             f"`share` is the fraction of the {POP_LABEL.get(pop, pop)} 1000 Genomes reference population expected to carry the same genotype "
+             f"(Hardy-Weinberg from the effect-allele frequency); the TSV has all five super-populations.\n")
     titles = dict(ancestry="Ancestry-informative markers", trait="Traits", health="Health-associated common variants",
                   carrier="Carrier / rare pathogenic sites typed on arrays", pgx="Pharmacogenomics")
     for cat in CATEGORY_ORDER:
@@ -112,7 +115,10 @@ def write(out: Path, loaded, qc: dict, chr_table: pd.DataFrame, roh: pd.DataFram
         if sub.empty:
             continue
         L.append(f"\n### {titles.get(cat, cat)}\n")
-        L.append(_md_table(sub, ["rsid", "gene", "label", "genotype", "effect_count", "interpretation"]))
+        sub = sub.copy()
+        col = f"gf_{pop}"
+        sub["share"] = sub[col].map(lambda v: f"{v:.0%}" if isinstance(v, float) else "") if col in sub else ""
+        L.append(_md_table(sub, ["rsid", "gene", "label", "genotype", "effect_count", "share", "interpretation"]))
     cv = extra.get("clinvar") or {}
     L.append("\n## 9. ClinVar scan of every typed site\n")
     L.append(extra.get("clinvar_text", "not run"))

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/report_html.py

@@ -10,7 +10,7 @@ import pandas as pd
 
 from . import __version__
 from .context import lookup
-from .panel import CATEGORY_ORDER
+from .panel import CATEGORY_ORDER, POPS, POP_LABEL
 from .report import DISCLAIMER
 
 CAT_TITLES = dict(ancestry="Ancestry-informative markers", trait="Traits",
@@ -26,10 +26,10 @@ SENSITIVE = {"health", "carrier"}
 
 CSS = """
 :root{color-scheme:light;--bg:#fcfcfb;--surface:#ffffff;--line:#e6e4df;--ink:#0b0b0b;--ink-2:#52514e;--ink-3:#8a8781;
---accent:#2a78d6;--accent-soft:#cde2fb;--accent-ink:#184f95;--good:#0ca30c;--warn:#fab219;--serious:#ec835a;--critical:#d03b3b;
+--accent:#7c3aed;--accent-soft:#efe9fd;--accent-ink:#5b21b6;--good:#0ca30c;--warn:#fab219;--serious:#ec835a;--critical:#d03b3b;
 --radius:12px;--mono:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace;--sans:-apple-system,BlinkMacSystemFont,"Segoe UI",Inter,Roboto,Helvetica,Arial,sans-serif}
-@media (prefers-color-scheme:dark){:root:not([data-theme=light]){color-scheme:dark;--bg:#141413;--surface:#1a1a19;--line:#2e2e2c;--ink:#f0efec;--ink-2:#c3c2b7;--ink-3:#8f8e86;--accent:#3987e5;--accent-soft:#0d366b;--accent-ink:#9ec5f4}}
-:root[data-theme=dark]{color-scheme:dark;--bg:#141413;--surface:#1a1a19;--line:#2e2e2c;--ink:#f0efec;--ink-2:#c3c2b7;--ink-3:#8f8e86;--accent:#3987e5;--accent-soft:#0d366b;--accent-ink:#9ec5f4}
+@media (prefers-color-scheme:dark){:root:not([data-theme=light]){color-scheme:dark;--bg:#141413;--surface:#1a1a19;--line:#2e2e2c;--ink:#f0efec;--ink-2:#c3c2b7;--ink-3:#8f8e86;--accent:#a78bfa;--accent-soft:#2e1a5e;--accent-ink:#c4b5fd}}
+:root[data-theme=dark]{color-scheme:dark;--bg:#141413;--surface:#1a1a19;--line:#2e2e2c;--ink:#f0efec;--ink-2:#c3c2b7;--ink-3:#8f8e86;--accent:#a78bfa;--accent-soft:#2e1a5e;--accent-ink:#c4b5fd}
 *{box-sizing:border-box}html{-webkit-text-size-adjust:100%}
 body{margin:0;background:var(--bg);color:var(--ink);font:15px/1.55 var(--sans)}
 .wrap{max-width:1040px;margin:0 auto;padding:0 16px 64px}
@@ -60,6 +60,7 @@ tr:first-child td{border-top:0}
 td.gt,td.num{font-family:var(--mono);font-variant-numeric:tabular-nums;white-space:nowrap}
 .badge{display:inline-block;min-width:34px;text-align:center;padding:1px 8px;border-radius:999px;font-size:12.5px;font-family:var(--mono);border:1px solid var(--line);color:var(--ink-2)}
 .badge.n1{border-color:var(--accent);color:var(--accent-ink)}.badge.n2{background:var(--accent);border-color:var(--accent);color:#fff}
+.bar{fill:var(--accent)}
 .badge.na{color:var(--ink-3);border-style:dashed}
 .readings{display:grid;grid-template-columns:repeat(auto-fit,minmax(260px,1fr));gap:12px}
 .readings .card .value{font-size:18px;font-weight:600;margin:2px 0 4px}.readings .card .note{font-size:13.5px;color:var(--ink-2)}
@@ -141,8 +142,16 @@ def _het_chart(ct: pd.DataFrame) -> str:
         'X near zero with Y present indicates a male sample; MT is always homozygous.</p>'
 
 
-def _panel_table(sub: pd.DataFrame) -> str:
-    h = ['<div class="tablewrap"><table class="panel"><thead><tr><th>Marker</th><th>Gene</th><th>What it tags</th><th>Genotype</th><th>Effect allele</th><th>Reading</th></tr></thead><tbody>']
+def _share_cell(r, pop: str) -> str:
+    v = r.get(f"gf_{pop}")
+    if not isinstance(v, float):
+        return '<td class="num"><span class="caption">—</span></td>'
+    parts = [f"{POP_LABEL[p]}: {r.get(f'gf_{p}'):.0%}" for p in POPS if isinstance(r.get(f"gf_{p}"), float)]
+    return f'<td class="num" title="{e("; ".join(parts))}">{v:.0%}</td>'
+
+
+def _panel_table(sub: pd.DataFrame, pop: str = "EUR") -> str:
+    h = [f'<div class="tablewrap"><table class="panel"><thead><tr><th>Marker</th><th>Gene</th><th>What it tags</th><th>Genotype</th><th>Effect allele</th><th title="Share of the {e(POP_LABEL.get(pop, pop))} 1000 Genomes reference population expected to carry the same genotype. Hover for all five super-populations.">How common ({e(pop)})</th><th>Reading</th></tr></thead><tbody>']
     for _, r in sub.iterrows():
         gt = r["genotype"]
         ec = str(r["effect_count"])
@@ -155,7 +164,7 @@ def _panel_table(sub: pd.DataFrame) -> str:
             gtc = e(gt)
         interp = e(str(r["interpretation"]))
         note = e(str(r.get("note", "") or ""))
-        h.append(f'<tr><td class="num">{e(r["rsid"])}</td><td>{e(r["gene"])}</td><td>{e(r["label"])}</td><td class="gt">{gtc}</td><td>{badge}</td><td>{interp}' + (f'<div class="caption">{note}</div>' if note else "") + "</td></tr>")
+        h.append(f'<tr><td class="num">{e(r["rsid"])}</td><td>{e(r["gene"])}</td><td>{e(r["label"])}</td><td class="gt">{gtc}</td><td>{badge}</td>{_share_cell(r, pop)}<td>{interp}' + (f'<div class="caption">{note}</div>' if note else "") + "</td></tr>")
     h.append("</tbody></table></div>")
     return "".join(h)
 
@@ -271,13 +280,14 @@ def write_html(out: Path, loaded, qc: dict, ct: pd.DataFrame, roh: pd.DataFrame,
                  f'<div class="readings"><div class="card"><h3>APOE</h3><div class="value">{e(v)}</div><div class="note">{e(n)}</div></div></div></details>')
     H.append("</section>")
     # Panel
-    H.append('<section id="panel"><h2>Marker panel</h2><p class="lead">Each row is one well-studied site. The badge counts how many copies of the <em>effect allele</em> you carry (0, 1 or 2) on the GRCh37 forward strand. "Not on array" means this vendor\'s chip does not type the site.</p>'
+    pop = extra.get("pop") or "EUR"
+    H.append(f'<section id="panel"><h2>Marker panel</h2><p class="lead">Each row is one well-studied site. The badge counts how many copies of the <em>effect allele</em> you carry (0, 1 or 2) on the GRCh37 forward strand. The share column says what fraction of the {e(POP_LABEL.get(pop, pop))} reference population (1000 Genomes) is expected to carry the same genotype; hover it for all five world regions. "Not on array" means this vendor\'s chip does not type the site.</p>'
              '<div class="filter"><input id="q" type="search" placeholder="Filter by gene, rsID or keyword…" aria-label="Filter markers"></div>')
     for cat in CATEGORY_ORDER:
         sub = panel[panel.category == cat]
         if sub.empty:
             continue
-        block = f'<div class="catblock"><h3>{e(CAT_TITLES.get(cat, cat))}</h3><p class="lead">{e(CAT_BLURB.get(cat, ""))}</p>{_panel_table(sub)}</div>'
+        block = f'<div class="catblock"><h3>{e(CAT_TITLES.get(cat, cat))}</h3><p class="lead">{e(CAT_BLURB.get(cat, ""))}</p>{_panel_table(sub, pop)}</div>'
         if cat in SENSITIVE:
             H.append(f'<details class="sensitive"><summary>{e(CAT_TITLES[cat])}<small>Collapsed on purpose: contains disease-risk and carrier results. Open when you are ready, or use "Reveal health sections" above.</small></summary>{block}</details>')
         else:

{gnomancer-0.3.2 → gnomancer-0.4.0}/gnomancer/summary.py

@@ -76,14 +76,17 @@ def at_a_glance(loaded, qc, roh, roh_text, combos, y, mt, ancestry, panel: pd.Da
         for _, r in p[p.category.isin(["health", "carrier"])].iterrows():
             n = str(r.effect_count).split("/")[0]
             if n == "2" and r.category == "health":
-                notable.append(f"{r.gene} {r.label}: {r.interpretation}.")
+                if "protective" in str(r.interpretation).lower():
+                    reassuring.append(f"{r.gene}: {r.interpretation}.")
+                else:
+                    notable.append(f"{r.gene} {r.label}: {r.interpretation}.")
             elif n == "1" and r.category == "carrier":
                 notable.append(f"{r.gene} {r.label}: carrier ({r.interpretation}).")
             elif n == "2" and r.category == "carrier":
                 notable.append(f"{r.gene} {r.label}: two copies ({r.interpretation}).")
         hets = p[(p.category == "health") & (p.effect_count.astype(str).str.startswith("1")) & ~p.interpretation.str.contains("benign", case=False)]
         if not hets.empty:
-            notable.append("One copy of a common risk allele at: " + ", ".join(f"{r.gene} ({r.label})" for _, r in hets.iterrows()) + ". Population-average territory, listed for completeness.")
+            notable.append("One copy of a risk allele at: " + ", ".join(f"{r.gene} ({r.label})" for _, r in hets.iterrows()) + ". Population-average territory, listed for completeness.")
     if "APOE" in combo:
         v = combo["APOE"][0].split(" (")[0]
         (reassuring if v == "e3/e3" else notable).append(f"APOE {v}: {combo['APOE'][1]}.")