Drop your 23andMe, AncestryDNA, or MyHeritage file and see what your DNA says about your sleep, caffeine, fitness, mood, and nutrition. Then plan with your own AI on top of it: what to eat, how to train, what to test next, built on peer-reviewed findings
No account, no card. Parsed on your device in about a minute.
Mainstream wellness guidance is built on population means. Your variants say where you sit relative to that mean. Five examples of what changes, with the genotype doing the work.
Of 1,023 adults in the study, ADORA2A rs2298383 T/T carriers reported poor sleep more often than C/C carriers. Same sleep questionnaire, one marker, a ten-point gap.
ADORA2A rs2298383 T/T is the version of the caffeine receptor gene whose carriers report poor sleep more often. The same study also measured dose: above 300 mg of caffeine a day, about three cups, sleep got shorter for every genotype. If you carry T/T, the cup count is the lever, and the generic two to three sits right at that line.
Each card tells you which way your genotype leans, how much it matters, and which study says so, across all five areas: body, nutrition, sleep, mind and longevity.
Your markers lean a little high, so it's worth knowing your numbers and checking now and then. Salt, weight, sleep and stress are the main levers.
Two G copies, no A allele. In East Asian populations the A allele tracks with slightly lower blood pressure, partly because A causes the alcohol-flush reaction so carriers drink less, though a more direct vascular effect is debated; G/G carries none of that flush-linked effect. A small, population-level pattern.
One G variant and one C baseline allele, a heterozygous C/G genotype. In research cohorts this genotype is linked to slightly higher measured blood pressure than the C/C baseline. This is a small, population-level shift.
Both copies at this position carry the T allele, a T/T genotype. In population studies the T allele is associated with modestly higher blood pressure, so a T/T reading sits slightly toward the higher-BP end of this single-marker range. This is a small leaning, not a diagnosis; the tiny push from one common variant is dwarfed by everyday factors like salt, weight, activity and the rest of your genome.
Both copies here carry the A allele, an A/A genotype, with no copies of the blood-pressure-raising G allele. In population studies measured blood pressure at this site sits toward the lower end of a very small range; each G copy is associated with only a tiny upward nudge.
The report connects to your AI over MCP, or travels as a bundle of plain files: your findings, your genotypes, and the instructions that teach an AI to use them. Either way, "what should I eat" and "why do I sleep badly" are answered from your genotype, with the numbers computed beforehand, not guessed by the chatbot.
instructions.md as your first message.Three findings in your report shape the answer:
Try: wake by 9am, sunlight and water first, omega-rich breakfast, coffee at 10:30, deep work 11 to 2, second block 3 to 5pm.
Want a few omega-rich breakfast recipes?
The same fixed pipeline for every file, with the real numbers at every step. The full write-up, including everything we refuse to claim, is public at Methods & limits.
parsed in your browser, the file stays with you
only what the knowledge base tracks, scoped to you
oriented and weighted by published effect sizes
verified against the record, audited by hand
The raw file is read in your browser and never uploaded. We only ever receive the ~600 markers your report needs, scoped to your account, deletable in one click.
The raw file is read in your browser and never uploaded. Only the markers your report needs reach our servers.
Wipe the markers and report we hold any time. No archives, no soft-delete.
The bundle lives on your device. Your AI conversations never pass through us.
We never feed your data into model training or sell datasets downstream, anonymized or otherwise.
Read the full Privacy Policy and Terms. How we read your DNA, what we combine, and what we refuse to claim is written out in Methods & limits.
Helisoma translates peer-reviewed population research into a personal summary. The insights are statistical associations, tendencies, not certainties. Most effects are small nudges. Not intended to diagnose, treat, cure, or prevent any disease. This is not medical advice. Consult a qualified professional before making health, diet, supplement, or medication decisions.
Real findings from your genome, free, right on this page. The full report is a single payment. No subscription, no renewals.
Payments are securely processed by Stripe. Because you can preview your report for free before paying, purchases are final once unlocked, though your statutory rights are unaffected. See our Terms for full details.