Codex v2 analysis upgrades

README.md

@@ -6,9 +6,9 @@
 [![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE)
 [![Python 3.11+](https://img.shields.io/badge/python-3.11%2B-blue.svg)](https://www.python.org/downloads/)
 
-**OpenDNA** parses a 23andMe / AncestryDNA / MyHeritage raw DNA file against 8 curated SNP panels (cardiovascular, methylation, pharmacogenomics, athletic, dietary, HFE, cognition, stimulant sensitivity), joins findings with ClinVar + PharmGKB/CPIC annotations, and renders a self-contained HTML report. It now also scores match confidence, shows panel coverage / blind spots, and rolls multi-marker genes like APOE, HFE, MTHFR, CYP2C19, and warfarin PGx into composite calls. Optional BYOK LLM synthesis (Anthropic Claude or OpenAI GPT) adds a prose interpretation layer.
+**OpenDNA** parses a 23andMe / AncestryDNA / MyHeritage raw DNA file against 12 curated SNP panels (cardiovascular, methylation, pharmacogenomics, athletic, dietary, HFE, histamine, cognition, stimulant sensitivity, eye health, vitamin D, nicotine), joins findings with ClinVar + PharmGKB/CPIC annotations, and renders a self-contained HTML report. It now also scores match confidence, shows panel coverage / blind spots, and rolls multi-marker patterns like APOE, HFE, MTHFR, histamine DAO/HNMT, CYP2C19, warfarin PGx, statin PGx, DPYD, AMD susceptibility, vitamin D tendency, and nicotine dependence into composite calls. Optional BYOK LLM synthesis (Anthropic Claude or OpenAI GPT) adds a prose interpretation layer.
 
-**What leaves your machine:** nothing, by default. No account, no upload, no telemetry. The raw DNA file is parsed, analyzed, and rendered entirely offline. If — and only if — you paste an API key and opt in to AI synthesis, OpenDNA sends the filtered *findings* (rsid + gene + genotype + tier + short note; roughly 40–50 lines of structured text) to the provider you chose. Your raw DNA file is never transmitted. See the [Privacy](#privacy--what-leaves-your-machine) section for the full rules.
+**What leaves your machine:** nothing, by default. No account, no upload, no telemetry. The raw DNA file is parsed, analyzed, and rendered entirely offline. If — and only if — you paste an API key and opt in to AI synthesis, OpenDNA sends the filtered *findings* (rsid + gene + genotype + tier + short note; roughly a few dozen lines of structured text) to the provider you chose. Your raw DNA file is never transmitted. See the [Privacy](#privacy--what-leaves-your-machine) section for the full rules.
 
 ## Sample report
 
@@ -17,7 +17,7 @@
 **[→ View the live sample report](https://raw.githack.com/corbett3000/OpenDNA/main/examples/sample-report.html)** (renders the actual HTML in your browser)
 · [view source](examples/sample-report.html)
 
-The sample was generated from a real 23andMe-style raw DNA file run through the full OpenDNA pipeline (parse → 8 panels → ClinVar + PharmGKB annotation → rule-based render). No personal raw data is embedded — only the interpreted findings.
+The sample was generated from a real 23andMe-style raw DNA file run through the full OpenDNA pipeline (parse → shipped panels → ClinVar + PharmGKB annotation → rule-based render). No personal raw data is embedded — only the interpreted findings.
 
 ---
 
@@ -32,7 +32,7 @@ OpenDNA is an educational and exploratory tool. **It is not a clinical diagnosti
 This is the whole trust claim; read it carefully:
 
 1. **Your raw DNA file never leaves your machine.** Parsing, panel matching, ClinVar + PharmGKB annotation, and the rule-based report all happen on `localhost`. No step in that pipeline touches the network.
-2. **LLM synthesis is opt-in and sends a minimal payload.** If and only if you paste an API key and click Generate, OpenDNA sends the filtered findings (roughly 40–50 lines of structured text — rsid, gene, genotype, tier, note) to the provider you chose. Your raw DNA file is **never** sent.
+2. **LLM synthesis is opt-in and sends a minimal payload.** If and only if you paste an API key and click Generate, OpenDNA sends the filtered findings (roughly a few dozen lines of structured text — rsid, gene, genotype, tier, note) to the provider you chose. Your raw DNA file is **never** sent.
 3. **No telemetry, no accounts, no server-side persistence.** API keys are read from the request, held in memory for one call, then discarded. There is no database.
 4. **The server binds to `127.0.0.1` by default.** Only the machine running it can reach it. `--host 0.0.0.0` requires an explicit flag and prints a warning.
 5. **Browser-local convenience storage is opt-in and bounded.** The web UI can remember your DNA file path, LLM provider, model, and API key in the browser's `localStorage` (scoped to `http://localhost:8787` on that one browser profile). Nothing is written to disk by the server, committed to git, or transmitted anywhere. The **Remember** checkbox controls it; the **Clear saved values** button wipes it instantly.
@@ -86,7 +86,7 @@ opendna --version
 pytest -v                           # (only if you installed [dev])
 ```
 
-Current contributor baseline: `51` tests passing.
+Current contributor baseline: `53` tests passing.
 
 ---
 
@@ -98,7 +98,7 @@ opendna serve
 ```
 
 1. **Paste the absolute path** to your raw DNA file (e.g. `/Users/you/Downloads/genome_yourname.txt`).
-2. **Pick panels** — all 8 are selected by default; uncheck to narrow the scope.
+2. **Pick panels** — all 12 are selected by default; uncheck to narrow the scope.
 3. **(Optional) AI synthesis** — pick Anthropic or OpenAI, pick a model, paste your API key. Leave the provider as *None* for a rule-based report only.
 4. **Click *Generate report*.** Watch the progress bar — you'll see the pipeline advance through validate → parse → analyze → annotate → LLM → render.
 5. **Download** the self-contained HTML or the structured JSON.
@@ -143,10 +143,14 @@ Every finding is tier-scored (`risk` / `warning` / `normal` / `unknown`), annota
 |---|---|---|
 | **Cardiovascular & Longevity** | CAD risk, thrombosis, lifespan | APOE, 9p21, FOXO3, LPA, F5, F2 |
 | **Methylation & Detox** | Folate/B12 cycle | MTHFR, COMT, MTR/MTRR, FUT2, CBS, VDR |
-| **Pharmacogenomics (PGx)** | Drug metabolism | CYP2C19, CYP2C9, VKORC1, CYP4F2, SLCO1B1, TPMT, CYP3A5 |
+| **Pharmacogenomics (PGx)** | Drug metabolism | CYP2C19, CYP2C9, VKORC1, CYP4F2, SLCO1B1, ABCG2, DPYD, TPMT, CYP3A5 |
 | **Athletic Performance & Recovery** | Fiber type, recovery | ACTN3, PPARA, PPARGC1A, COL5A1, IL6 |
 | **Dietary Sensitivity** | Lactose, caffeine, alcohol, omega-3 | LCT, CYP1A2, ALDH2, FADS1, TAS2R38 |
+| **Eye Health & Macular Degeneration** | Common AMD predisposition | CFH, ARMS2, C3 |
 | **Iron Metabolism (HFE)** | Hemochromatosis | C282Y, H63D, S65C |
+| **Histamine Handling** | Exploratory histamine breakdown | AOC1 (DAO), HNMT |
+| **Nicotine Dependence & Smoking Response** | Smoking heaviness, nicotine reinforcement | CHRNA5, CHRNA3 |
+| **Vitamin D & Bone** | Low-25(OH)D tendency, signaling | DHCR7, CYP2R1, VDR |
 | **Cognition & Mood** | Plasticity, dopamine | BDNF, DRD2, OXTR |
 | **Stimulant Sensitivity** | Caffeine/adenosine | ADORA2A, CYP1A2, COMT |
 
@@ -157,6 +161,8 @@ The analyzer automatically handles both **allele-order variations** (`AG` == `GA
 - `Not on chip` means the marker was not present in the source file. It does **not** mean the user has a reassuring genotype there.
 - `No-call` means the vendor included the marker but did not make a confident genotype call.
 - Composite calls are only made where the source file type can support them. OpenDNA still does **not** infer rare variants, structural variants, HLA types, CYP2D6 star alleles, or methylation from array data.
+- The histamine panel is exploratory. It can flag lower-clearance DAO/HNMT patterns, but it does **not** diagnose histamine intolerance, food reactions, or mast-cell disease on its own.
+- The AMD, vitamin D, and nicotine panels are predisposition panels. They do **not** diagnose current eye disease, vitamin D deficiency, or addiction by themselves.
 
 ---
 

docs/superpowers/handover-2026-04-17.md

@@ -4,7 +4,7 @@
 
 ## What this is
 
-Local-first web utility that parses a raw 23andMe-style DNA file, matches against 8 curated SNP panels, joins findings with ClinVar + PharmGKB/CPIC annotations, and renders a self-contained HTML report. Optional BYOK LLM synthesis (Anthropic default with prompt caching, OpenAI secondary) adds prose interpretation.
+Local-first web utility that parses a raw 23andMe-style DNA file, matches against 12 curated SNP panels, joins findings with ClinVar + PharmGKB/CPIC annotations, and renders a self-contained HTML report. Optional BYOK LLM synthesis (Anthropic default with prompt caching, OpenAI secondary) adds prose interpretation.
 
 - **Repo:** https://github.com/corbett3000/OpenDNA — **public**, Apache 2.0
 - **Runs at:** http://localhost:8787 (default)
@@ -16,16 +16,16 @@ Local-first web utility that parses a raw 23andMe-style DNA file, matches agains
 
 ### Shipped features
 
-- 8 curated SNP panels (cardiovascular, methylation, pharmacogenomics, athletic, dietary, HFE, cognition, stimulant sensitivity) — now expanded with F5 / F2 thrombophilia markers and CYP4F2 warfarin PGx support.
+- 12 curated SNP panels (cardiovascular, methylation, pharmacogenomics, athletic, dietary, HFE, histamine, cognition, stimulant sensitivity, eye health, vitamin D, nicotine) — now expanded with F5 / F2 thrombophilia markers, CYP4F2 / ABCG2 / DPYD PGx support, exploratory DAO / HNMT histamine handling coverage, common AMD susceptibility markers, vitamin D tendency markers, and nicotine-dependence markers.
 - ClinVar (16 rsids) + PharmGKB (8 rsids, CPIC drug recs for TPMT / CYP2C19 / CYP2C9 / VKORC1 / CYP4F2 / SLCO1B1) shipped subsets.
 - FastAPI server with five endpoints: `/`, `/api/panels`, `/api/analyze`, `/api/analyze-stream` (SSE for live progress), `/api/update-db`.
 - Vanilla-JS SPA: DOM APIs only (no `innerHTML` on dynamic data), report rendered into sandboxed iframe. Live progress bar with per-stage messages. Optional browser `localStorage` persistence of form values behind a *Remember* checkbox + *Clear saved values* button.
 - CLI: `opendna serve`, `opendna scan <file>`, `opendna update-db`.
 - Jinja2 HTML report (autoescape on) styled after the original sandbox synthesis — stat tiles, color-coded card grid per panel, critical-findings alert box for risk tier, AI synthesis block when LLM is configured.
-- Report now includes source-file QC, vendor/build heuristics, panel completeness, confidence labels, gene caveats, and composite summaries for APOE, HFE, MTHFR, CYP2C19, and warfarin PGx.
+- Report now includes source-file QC, vendor/build heuristics, panel completeness, confidence labels, gene caveats, and composite summaries for APOE, HFE, MTHFR, histamine DAO/HNMT, CYP2C19, warfarin PGx, statin PGx, DPYD, AMD susceptibility, vitamin D tendency, and nicotine dependence.
 - LLM provider layer: pluggable ABC, Anthropic default (`claude-sonnet-4-6` + prompt caching), OpenAI (`gpt-4o` using `max_completion_tokens`). API keys held per-request in memory only.
 - Analyzer normalizes allele order (`AG == GA`) AND reverse complement (panel `CC` matches file `GG` for non-palindromic C/T SNPs).
-- 51 tests passing (parser, panels, analyzer, summaries, annotations, report, LLM providers, server, CLI, SSE stream, e2e). `ruff check` clean. CI runs on Python 3.11 / 3.12 / 3.13.
+- 53 tests passing (parser, panels, analyzer, summaries, annotations, report, LLM providers, server, CLI, SSE stream, e2e). `ruff check` clean. CI runs on Python 3.11 / 3.12 / 3.13.
 - Apache 2.0 license.
 - README with full setup guide, per-provider raw-DNA download instructions, sample report + screenshot, troubleshooting, author links (`@corbett3000`, `stillrush.co/bio`).
 - Sample report: `examples/sample-report.html` (served live via raw.githack; preview image at `docs/assets/sample-report.png`).
@@ -55,12 +55,10 @@ If Claude Code opens this repo, `CLAUDE.md` at the root will auto-load with a co
 ```bash
 cd ~/Coding/OpenDNA
 source .venv/bin/activate          # or create one: python3 -m venv .venv && pip install -e ".[dev,all]"
-pytest -v                          # 44 tests should pass
+pytest -v                          # 53 tests should pass
 opendna serve                      # localhost:8787
 ```
 
-The current baseline is `51` passing tests.
-
 ## Key design decisions (for context)
 
 | Decision | Choice | Why |

src/opendna/panels/eye_health.json

@@ -0,0 +1,40 @@
+{
+  "id": "eye_health",
+  "name": "Eye Health & Macular Degeneration",
+  "description": "Common age-related macular degeneration (AMD) susceptibility markers. Best read as predisposition, not diagnosis.",
+  "snps": [
+    {
+      "rsid": "rs1061170",
+      "gene": "CFH",
+      "variant_name": "Y402H",
+      "description": "Complement-factor H variant strongly associated with common AMD risk in European-ancestry studies.",
+      "interpretations": {
+        "TT": {"tier": "normal", "note": "No common CFH Y402H risk allele detected at this site."},
+        "CT": {"tier": "warning", "note": "One common AMD risk allele is present at CFH. Lifestyle and retinal screening matter more as age advances."},
+        "CC": {"tier": "risk", "note": "Two common CFH risk alleles are present. This is a stronger inherited AMD predisposition signal."}
+      }
+    },
+    {
+      "rsid": "rs10490924",
+      "gene": "ARMS2",
+      "variant_name": "A69S",
+      "description": "One of the best-replicated common AMD variants; smoking amplifies the risk signal.",
+      "interpretations": {
+        "GG": {"tier": "normal", "note": "No common ARMS2 risk allele detected at this site."},
+        "GT": {"tier": "warning", "note": "One common AMD risk allele is present at ARMS2."},
+        "TT": {"tier": "risk", "note": "Two ARMS2 risk alleles are present. This is a strong common susceptibility signal for AMD."}
+      }
+    },
+    {
+      "rsid": "rs2230199",
+      "gene": "C3",
+      "variant_name": "R102G",
+      "description": "Complement C3 variant associated with advanced AMD risk in multiple cohorts.",
+      "interpretations": {
+        "CC": {"tier": "normal", "note": "No common C3 AMD risk allele detected at this site."},
+        "CG": {"tier": "warning", "note": "One C3 risk allele is present. This modestly raises common AMD susceptibility."},
+        "GG": {"tier": "risk", "note": "Two C3 risk alleles are present. This is a stronger complement-pathway AMD signal."}
+      }
+    }
+  ]
+}

src/opendna/panels/histamine.json

@@ -0,0 +1,62 @@
+{
+  "id": "histamine",
+  "name": "Histamine Handling",
+  "description": "Exploratory panel for histamine breakdown via DAO (AOC1) and HNMT. Useful for pattern-spotting, not for diagnosing histamine intolerance or mast-cell disease.",
+  "snps": [
+    {
+      "rsid": "rs10156191",
+      "gene": "AOC1",
+      "variant_name": "DAO Thr16Met",
+      "description": "Coding variant in the DAO enzyme. The T allele has been associated with reduced DAO activity in human association studies.",
+      "interpretations": {
+        "CC": {"tier": "normal", "note": "No reduced-DAO allele at this site."},
+        "CT": {"tier": "warning", "note": "One reduced-DAO allele. Histamine clearance may be a bit less resilient, especially with high histamine load or gut inflammation."},
+        "TT": {"tier": "risk", "note": "Two reduced-DAO alleles at this site. This is a stronger genetic signal for lower DAO-mediated histamine breakdown."}
+      }
+    },
+    {
+      "rsid": "rs1049742",
+      "gene": "AOC1",
+      "variant_name": "DAO Ser332Phe",
+      "description": "DAO coding variant associated with lower enzyme activity in AOC1 deficiency studies.",
+      "interpretations": {
+        "CC": {"tier": "normal", "note": "No reduced-DAO allele at this site."},
+        "CT": {"tier": "warning", "note": "One reduced-DAO allele. On its own this is usually a mild signal, but it matters more when other DAO variants stack up."},
+        "TT": {"tier": "risk", "note": "Two reduced-DAO alleles at this site. Stronger genetic signal for lower DAO function."}
+      }
+    },
+    {
+      "rsid": "rs1049793",
+      "gene": "AOC1",
+      "variant_name": "DAO His664Asp",
+      "description": "DAO coding variant associated with reduced histamine-degrading capacity in AOC1 studies.",
+      "interpretations": {
+        "CC": {"tier": "normal", "note": "No reduced-DAO allele at this site."},
+        "CG": {"tier": "warning", "note": "One reduced-DAO allele. This can contribute to lower histamine tolerance when paired with other DAO variants."},
+        "GG": {"tier": "risk", "note": "Two reduced-DAO alleles at this site. Stronger signal for lower DAO-mediated histamine breakdown."}
+      }
+    },
+    {
+      "rsid": "rs2052129",
+      "gene": "AOC1",
+      "variant_name": "DAO promoter c.-691G>T",
+      "description": "Promoter-region variant associated with lower AOC1 transcription and lower DAO activity.",
+      "interpretations": {
+        "GG": {"tier": "normal", "note": "No lower-expression DAO allele at this site."},
+        "GT": {"tier": "warning", "note": "One lower-expression allele. DAO reserve may be somewhat lower under stressors like alcohol, infection, or gut irritation."},
+        "TT": {"tier": "risk", "note": "Two lower-expression alleles at this site. Stronger signal for reduced DAO expression."}
+      }
+    },
+    {
+      "rsid": "rs11558538",
+      "gene": "HNMT",
+      "variant_name": "HNMT Thr105Ile",
+      "description": "Functional HNMT variant. The T allele lowers histamine N-methyltransferase activity and can reduce intracellular histamine clearance.",
+      "interpretations": {
+        "CC": {"tier": "normal", "note": "Standard HNMT activity at this site."},
+        "CT": {"tier": "warning", "note": "One lower-activity HNMT allele. Histamine breakdown outside the gut may be somewhat slower."},
+        "TT": {"tier": "risk", "note": "Two lower-activity HNMT alleles. Stronger signal for reduced HNMT-mediated histamine breakdown."}
+      }
+    }
+  ]
+}

src/opendna/panels/nicotine.json

@@ -0,0 +1,40 @@
+{
+  "id": "nicotine",
+  "name": "Nicotine Dependence & Smoking Response",
+  "description": "Common smoking-heaviness and nicotine-reinforcement markers. Behavioral predisposition only; not deterministic.",
+  "snps": [
+    {
+      "rsid": "rs16969968",
+      "gene": "CHRNA5",
+      "variant_name": "D398N",
+      "description": "The strongest common CHRNA5 smoking-dependence marker in many GWAS; associated with heavier smoking among smokers.",
+      "interpretations": {
+        "GG": {"tier": "normal", "note": "No common CHRNA5 nicotine-risk allele detected at this site."},
+        "AG": {"tier": "warning", "note": "One CHRNA5 risk allele is present. If nicotine use starts, reinforcement can run stronger than baseline."},
+        "AA": {"tier": "risk", "note": "Two CHRNA5 risk alleles are present. This is the strongest common dependence-prone pattern at this site."}
+      }
+    },
+    {
+      "rsid": "rs1051730",
+      "gene": "CHRNA3",
+      "variant_name": "Smoking quantity marker",
+      "description": "Closely related to smoking heaviness and toxin exposure among smokers.",
+      "interpretations": {
+        "GG": {"tier": "normal", "note": "No common CHRNA3 smoking-quantity risk allele detected at this site."},
+        "AG": {"tier": "warning", "note": "One CHRNA3 risk allele is present. This can track with heavier smoking once nicotine use begins."},
+        "AA": {"tier": "risk", "note": "Two CHRNA3 risk alleles are present. This is a stronger common signal for heavier smoking exposure."}
+      }
+    },
+    {
+      "rsid": "rs588765",
+      "gene": "CHRNA5",
+      "variant_name": "Regulatory expression marker",
+      "description": "Expression-linked CHRNA5 marker that refines nicotine-dependence tendency when paired with rs16969968.",
+      "interpretations": {
+        "TT": {"tier": "normal", "note": "No common rs588765 risk allele detected at this site."},
+        "CT": {"tier": "warning", "note": "One regulatory risk allele is present. This is a smaller signal than rs16969968 but can reinforce it."},
+        "CC": {"tier": "warning", "note": "Two regulatory risk alleles are present. This is a moderate added nicotine-reinforcement signal."}
+      }
+    }
+  ]
+}