Analyze your genetic data through conversation with Claude. Works with 23andMe, AncestryDNA, MyHeritage, Nebula, and other DNA testing services. Run scripts, explore your genome, and get personalized explanations — all in natural dialogue.
Traditional DNA analysis tools give you static reports. With Claude, you get:
- Interactive exploration — Ask follow-up questions about any result
- Plain English explanations — No need to Google every SNP
- Cross-category insights — Connect findings across health, nutrition, ancestry
- Personalized context — Discuss how results apply to your specific situation
- Research deep-dives — Ask Claude to explain the science behind any marker
- Export raw data from your DNA testing service (see Supported Services)
- Place the file in
data/folder - Open Claude Code in this project directory
- Tell Claude: "Update scripts to use my genome file: data/your_filename.txt"
- Start chatting:
You: Run the health analysis script
Claude: *runs health_analysis.py, shows results*
You: What does my MTHFR status mean for daily life?
Claude: *explains your specific genotype and practical implications*
You: Should I be concerned about the APOE result?
Claude: *provides context, explains penetrance, suggests next steps*
| Script | What it analyzes |
|---|---|
health_analysis.py |
Cardiovascular, oncology, neurology, diabetes, pharmacogenomics |
ancestry_analysis.py |
Ethnic markers, haplogroups, pigmentation genes |
nutrition_analysis.py |
Lactose, gluten, caffeine, alcohol, vitamins metabolism |
carrier_status_analysis.py |
Hereditary condition carrier status |
cognitive_analysis.py |
Cognitive traits and predispositions |
longevity_analysis.py |
Longevity-associated variants |
psychology_analysis.py |
Psychological traits |
sports_fitness_analysis.py |
Athletic performance genetics |
sleep_chronotype_analysis.py |
Sleep patterns, circadian rhythm |
immunity_analysis.py |
Immune system markers |
detoxification_analysis.py |
Detox pathways, enzyme activity |
skin_analysis.py |
Skin characteristics |
vision_hearing_analysis.py |
Sensory genetics |
pain_sensitivity_analysis.py |
Pain perception |
reproductive_analysis.py |
Reproductive health |
physical_traits_analysis.py |
Physical characteristics |
You: Run health analysis and tell me the most important findings
You: I see I have a variant in CYP2D6. What medications should I be careful with?
You: My doctor wants to prescribe codeine. Is that safe for my genotype?
You: Analyze my ancestry markers
You: What does the EDAR variant tell us about my ancestry?
You: How do my haplogroups connect to migration patterns?
You: What does my genome say about coffee metabolism?
You: Based on my variants, what supplements might actually help me?
You: I'm lactose intolerant symptoms-wise. What does my genetics show?
You: Explain the APOE gene and why it matters
You: What's the current research on rs1801133?
You: How reliable are these SNP associations? What's the evidence quality?
You: Run all analysis scripts
Claude: *runs 17 analysis scripts, generates reports/*
You: Generate DNA Terminal webpage
Claude: *reads reports, creates webpage/dna_terminal.html*
You: Open the webpage in browser
Claude: *opens file in default browser*
You: Run complete DNA analysis pipeline — all scripts, then generate webpage
dna-claude-analysis/
├── data/
│ ├── README.md # Instructions for data placement
│ └── your_genome.txt # ← Put your DNA file here
├── scripts/
│ └── *_analysis.py # Analysis scripts (17 categories)
├── reports/ # Generated markdown reports
├── webpage/
│ ├── STYLE_GUIDE.md # Terminal UI design system
│ └── dna_terminal.html # ← Generated webpage
├── .claude/ # Claude project settings
├── CLAUDE.md # Project instructions for Claude
├── .gitignore # Keeps your data private
└── README.md
Important: Before running scripts, you need to set the path to your genome file.
Just tell Claude:
You: Update all scripts to use my genome file: data/my_dna_file.txt
Claude will update the GENOME_FILE variable in all scripts automatically.
Each script has a GENOME_FILE variable at the top:
# In scripts/health_analysis.py (and other scripts)
GENOME_FILE = f"{BASE_PATH}/data/your_genome_file.txt" # ← Change thisReplace with your actual filename:
GENOME_FILE = f"{BASE_PATH}/data/genome_John_Doe_v5_Full_20240101.txt"┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ Your Genome │ ──▶ │ Python Script │ ──▶ │ Markdown Report│
│ (600K SNPs) │ │ (analyze SNPs) │ │ (findings) │
└─────────────────┘ └─────────────────┘ └─────────────────┘
│
┌───────────────────┴───────────────────┐
▼ ▼
┌─────────────────┐ ┌─────────────────┐
│ Claude │ │ DNA TERMINAL │
│ (explain, │ │ (interactive │
│ contextualize,│ │ webpage) │
│ answer Qs) │ │ │
└─────────────────┘ └─────────────────┘
After running analysis scripts, Claude can generate a beautiful DNA Terminal webpage — a single-file HTML visualization with terminal/hacker aesthetic.
You: Generate DNA Terminal webpage from the reports
Claude: *reads all reports, creates webpage/dna_terminal.html*
- Single HTML file — No dependencies, works offline
- Terminal aesthetic — Matrix-style green-on-black interface
- Fixed navigation — Jump between sections
- Color-coded risks — Green (good), amber (warning), red (risk)
- Mobile responsive — Works on any device
See webpage/STYLE_GUIDE.md for complete design system:
- Color palette and typography
- Component library (tables, charts, findings)
- Section structure and IDs
- Build instructions
┌─────────────────────────────────────────────────────────────────────┐
│ DNA ANALYSIS PIPELINE │
├─────────────────────────────────────────────────────────────────────┤
│ │
│ STEP 1: DATA │
│ ───────────────── │
│ • Export raw DNA from 23andMe/Ancestry/etc │
│ • Place in data/ folder │
│ • Tell Claude: "Use my file: data/genome.txt" │
│ │
│ STEP 2: ANALYSIS │
│ ───────────────── │
│ • "Run health analysis" — single script │
│ • "Run all analysis scripts" — full pipeline │
│ • Reports saved to reports/*.md │
│ │
│ STEP 3: EXPLORE │
│ ───────────────── │
│ • Ask questions about any finding │
│ • "What does my APOE status mean?" │
│ • "How does this affect my diet?" │
│ │
│ STEP 4: VISUALIZE │
│ ───────────────── │
│ • "Generate DNA Terminal webpage" │
│ • Opens interactive HTML dashboard │
│ • Terminal aesthetic with navigation │
│ │
└─────────────────────────────────────────────────────────────────────┘
- Start broad, then narrow — Run a full analysis first, then ask about specific findings
- Ask "so what?" — Claude can explain practical implications, not just raw data
- Request comparisons — "How does my result compare to population averages?"
- Explore connections — "Could my sleep issues be related to any of these variants?"
- Verify important findings — Ask Claude about evidence quality and study sizes
- Not medical advice — This is educational exploration, not diagnosis
- Genetics ≠ destiny — Environment and lifestyle matter more for most traits
- Consult professionals — For medical decisions, see a genetic counselor
- Privacy first — Your genome data stays local on your machine
- Python 3.6+
- Claude Code (or any Claude interface with file access)
- Raw DNA data from any major testing service
Scripts work with raw data from multiple providers. Most use similar formats (rsID, chromosome, position, genotype).
| Service | Coverage | Raw Data Export | Notes |
|---|---|---|---|
| 23andMe | ~640K SNPs | Settings → 23andMe Data → Download | Most popular, good health reports |
| AncestryDNA | ~700K SNPs | Settings → Download DNA Data | Best for genealogy |
| MyHeritage DNA | ~700K SNPs | DNA → Manage DNA kits → Download | Good European coverage |
| FamilyTreeDNA | ~700K SNPs | myFTDNA → Download Raw Data | Y-DNA & mtDNA specialists |
| Living DNA | ~650K SNPs | Account → Download | Best British Isles ancestry |
| Service | Coverage | Price Range | Notes |
|---|---|---|---|
| Nebula Genomics | 30x WGS | $250-300 | Privacy-focused, good reports |
| Dante Labs | 30x WGS | $200-400 | European company |
| Sequencing.com | 30x WGS | $200-400 | Good app ecosystem |
| Sano Genetics | 30x WGS | ~$300 | UK-based |
| Service | Type | Notes |
|---|---|---|
| All of Us | WGS | NIH research program (US only) |
| Our Future Health | Genotyping | UK research program |
| Genes for Good | Genotyping | University of Michigan (limited) |
Already tested? Upload your raw data for additional analysis:
- Promethease — $12, comprehensive health report
- GEDmatch — Free, genealogy & ethnicity
- Genetic Genie — Free, methylation & detox
- Codegen.eu — Free, various health reports
- DNA.Land — Free, ancestry & traits
- Log into 23andMe
- Settings → 23andMe Data → Download Raw Data
- Unzip and place in
data/folder
- Log into Ancestry
- DNA → Settings → Download DNA Data
- Confirm via email, download, place in
data/
- Log into MyHeritage
- DNA → Manage DNA kits → Actions → Download
- Place in
data/folder
Most services: Account Settings → Download Raw Data → Place in data/
Scripts expect standard format (auto-detected):
# rsid chromosome position genotype
rs4477212 1 82154 AA
rs3094315 1 752566 AG
Most services use this format. VCF files from WGS may need conversion (ask Claude for help).
Explore your genome through conversation. Ask Claude anything.