Your DNA already knows how you'll respond to medications. We translate it — in 15 seconds, for $49.
Upload the raw 23andMe, AncestryDNA, or MyHeritage file you already have. You'll get a personalized drug-response report covering 74+ medications graded for your specific genome — plus a chat you can ask anything, and a citation-bearing PDF to hand to your prescriber. No new DNA test required, no doctor's order required, no two-week wait.
Works with your existing raw .txt, .csv, or .zip download. Processing takes ~15 seconds.
What a real finding looks like
For someone with CYP2C19 *2/*2 (a poor metabolizer phenotype, present in ~3% of people), PharmTwin flags concrete recommendations across categories:
Avoid Clopidogrel (Plavix) — significantly reduced active-metabolite formation; CPIC recommends prasugrel or ticagrelor instead.
Adjust Escitalopram (Lexapro), Sertraline (Zoloft), Citalopram (Celexa) — 50% dose reduction recommended; higher plasma concentrations increase side-effect risk.
Adjust Omeprazole, Pantoprazole, Lansoprazole, Esomeprazole (Nexium) — consider 50% reduction on chronic therapy; higher exposure.
Avoid Tertiary-amine tricyclics (amitriptyline, imipramine, doxepin) — sub-optimal response; safer secondary amines available.
Standard 5-FU, capecitabine, the thiopurines, the G6PD-sensitive drugs — no genotype-guided adjustment needed.
That's a slice of one phenotype. The full report covers 74 medications across 23 pharmacogenes for your specific genome, every recommendation traceable to CPIC, DPWG, or FDA source text.
What you get
Personalized drug-response report
Every CPIC-graded medication classified for your genome as Avoid, Adjust, or Standard, organized by category (mental health, cardiovascular, GI, anesthesia, oncology, etc.).
Plain-English explanations
For every flagged drug, a 2-3 sentence explanation written specifically for your phenotype — what it means and what to bring up with your prescriber. No jargon.
Conversational chat
Ask anything in natural language ("Is Lexapro right for me?", "Anything I should tell my anesthesiologist?"). Grounded in your actual variants — refuses to invent.
Doctor-handoff PDF
A clinician-grade printable PDF with verbatim CPIC text, PubMed citations, and an explicit "this is consumer data, confirm with CLIA testing" disclaimer. Print it, hand it to your prescriber.
Multi-source confirmation
We pull from CPIC, DPWG (Dutch PGx), the FDA's drug-label annotations, and the FDA's Pharmacogenomic Associations table. Every drug card shows which sources concur — so you can see the difference between "CPIC Strong + DPWG + FDA all say avoid" and "one FDA label mentions this."
Two-pass CYP2D6 calling
Most consumer pharmacogenomic products skip CYP2D6 from SNP-array data. We run PharmCAT twice — once in research mode to call CYP2D6 SNVs, once in standard mode to apply that call to every CYP2D6-dependent drug recommendation. The same approach a clinical lab uses for non-CNV CYP2D6 variants.
Why this is different from clinical pharmacogenomic tests
| PharmTwin | GeneSight / Genomind | 23andMe PGx add-on | |
|---|---|---|---|
| Price | $49 | $300–600 | $30 (very limited) |
| Time to result | 15 seconds | 2 weeks | Minutes |
| Doctor's order required | No | Yes | No |
| Medications graded for your genome | 74+ actionable | ~60 | 8 |
| Conversational chat | Yes | No | No |
| Doctor-handoff PDF with PMID citations | Yes | Summary letter only | No |
| Multi-source attribution (CPIC + DPWG + FDA) | Yes, per drug | CPIC only | Limited |
| Uses your existing DNA file | Yes | No (their lab) | n/a (their data) |
| CLIA-certified | No (disclosed) | Yes | Yes |
We're explicit about the trade-off: PharmTwin runs on consumer SNP-array data, not CLIA-certified lab data. For any clinically actionable finding the doctor PDF says, in print, to confirm with CLIA-certified testing before changing therapy. That's how this should be used.
How it works, in three steps
- Upload your raw DNA file. The .txt, .csv, or .zip you downloaded from 23andMe, AncestryDNA, or MyHeritage. Nothing else required.
- We call your pharmacogene variants using PharmCAT (the open-source variant-caller from Stanford that clinical labs also use), then match them against CPIC, DPWG, and FDA drug-gene guidelines. Two-pass CYP2D6 enrichment included.
- You get your report + chat + downloadable doctor PDF, all at a private URL we email you. About 15 seconds end to end.