Upload your DNA file, blood panels and wearable data. The Whale DNA agent reads them, explains what matters, and answers your questions — cross-referenced against ClinVar and PharmGKB.
Hi — I'm the Whale DNA agent. I read the health data you already have — your DNA file, blood panels, wearable export — and turn it into something useful. Let's start with your genome. Drop your 23andMe / AncestryDNA / MyHeritage file below (paperclip ↓), or try a sample. Raw files are deleted within 24 hours.
Educational, not medical. Raw files deleted within 24h.
Most consumer DNA reports stop at "you might prefer cilantro." We tell you exactly which variant we found, what the literature says, and what to do tomorrow morning.
C677T homozygotes lose ~70% of the enzyme that converts folate. Standard supplements (folic acid in cereal, prenatal vitamins) sit unused or worse.
Poor metabolizers can't activate clopidogrel — the prodrug stays inert. PharmGKB recommends an alternative antiplatelet for these patients.
Heterozygotes express both alpha-actinin-3 and -2. Genetic studies show roughly balanced sprint/endurance potential rather than strong specialization either way.
Reports cite primary sources (ClinVar, PharmGKB, GWAS Catalog). Educational, not medical advice.
Paste a raw 23andMe file into a general chatbot and it pattern-matches from memory — it can't tell which of your ~600,000 positions matter, confuses risk alleles and DNA strands, and invents meanings for random SNPs. We never ask a model to interpret your DNA. The matching is deterministic code against real databases; the model only explains what's already verified.
We look across the whole karyotype — not just one panel. Every blue dot is a gene we scan for variants that change drug response, disease risk, or how your body handles nutrients.
Every report covers eight categories of findings. We label each one with its evidence level — so you know what your doctor would act on, what's published but personalize with judgment, and what's just fun to know.
Every cited finding links to its source paper or database entry. No hand-wavy "studies show."
You don't need a new test or another wearable. We read what you already have — and the more signals you bring, the sharper the report.
Your raw file holds half a million genotype calls. We filter, cross-reference and translate them into eight sections a human (and their doctor) can act on.
Genotype TSV streamed in-memory, never written to disk in raw form.
Only clinically-actionable loci leave our parser. The rest are discarded.
2.8M ClinVar entries · 730 drug-gene pairs · 200K GWAS associations.
Claude-authored under a strict template. Every claim cites its source.
Emailed to you. Saved to your account so you can regenerate with new data.
AWS-enforced S3 lifecycle deletes the original file before the next sunrise. We can't override it. The report stays — derived, anonymized, keyed to a SHA-256 pseudoID, not your name.
How the anonymization worksTry us with your genome only.
Full stack: genome + labs + wearables + symptoms.
Your body has been logging itself for years — across 23andMe, your last blood panel, your watch. We finally read it together, in one place.
Start your report