Where two models both fit, you get both
Ancestry is often genuinely ambiguous. When two models survive the same checks, your report carries both — the data may simply not separate them.
qpAdm ancestry decomposition
Upload the raw data file from 23andMe, AncestryDNA or MyHeritage. We fit it against published ancient reference populations and send back the model, the weights and the standard errors — not a pie chart.
US$19.99 · one report · no account, no subscription · results in 24–48 hours
Charged in US dollars. Your bank sets the final amount in your currency.
Punjabi Brahmin
Above 0.05, so this model is not rejected — which is not the same as proven. Each ± is one standard error on the weight beside it.
The tested sources, the weight on each, the uncertainty on every weight, and whether the model survives its own checks.
Ancestry is often genuinely ambiguous. When two models survive the same checks, your report carries both — the data may simply not separate them.
Each ± is one standard error on the weight beside it. Without an uncertainty estimate, there is no way to tell how precise the weight is.
If no tested model survives, the report says so. That is the completed analysis, not a failure to process your file.
The ancient populations behind your model are named in the report — not hidden behind a brand name for a region.
The .txt, .csv or .zip from your testing service. Nothing
to install, no account to create.
US$19.99 at a secure checkout — we never receive your card number. One report, no subscription, nothing recurring.
A private link by email within 24–48 hours — not minutes, because a person runs the model and checks whether it survives.
The .txt, .csv or .zip export from your
testing service. Up to 64 MB.
Don't have it yet? How to download your raw data →
The first three are tested against every release. Autosomal exports from other services usually work too. If yours cannot be processed, no analysis runs and you are refunded in full, automatically.
US$19.99 · one report · no account, no subscription.
If your file cannot be processed, you are refunded in full, automatically.
The raw autosomal export from your testing service — the .txt, .csv or .zip you download from your account, up to 64 MB. We test 23andMe, AncestryDNA and MyHeritage against every release; FamilyTreeDNA, Living DNA and most others usually work too. Not a PDF of your results, not a Y-only extract, and not an exome or clinical VCF. There is a step-by-step download guide linked above.
One report, emailed as a private link within 24–48 hours: the fitted model, the weight on each source population, the standard error on every weight, and the p-value for the fit. If competing models also fit, you get those too, because that is the honest answer.
Those usually compare you to modern reference panels and emphasise a single confident-looking percentage. qpAdm fits you against published ancient genomes, reports a standard error on every weight and a p-value for the fit, and can reject its own model. You get the uncertainty made explicit instead of a clean pie chart — less tidy, more honest about what the data can and cannot say.
It is uploaded over an encrypted connection, stored encrypted, used only to produce your report, and never sold. It is disclosed only to the service providers needed to run the service, as set out in the privacy policy — never to advertisers or data brokers. How long we keep it is set out in the terms of service. You can ask us to delete your copy at any time by emailing support@indoaryan.com.
Automatically, in full, if your file cannot be processed — a PDF instead of raw data, a Y-only extract, or too few usable markers. It is not refunded when the file processes fine and no model survives the checks: that is the completed analysis and it is your report. The distinction is whether we could run it, not whether you liked the answer.
No. It is a research-grade population-genetics analysis. It is not medical, diagnostic, paternity or forensic, and it establishes no nationality, community, caste or entitlement.
“The p-value and the ± on every weight is what I wanted. My 23andMe pie chart never told me how sure it was.”
Priya S. · Bengaluru
“Two models both fit and the report said so instead of picking one. That honesty is why I trust the numbers.”
Arjun M. · Toronto
“Took about 30 hours. Clear PDF, named every ancient population used. Would like a plain-language page one.”
Neha K. · Pune
“Finally a service that runs real qpAdm on my raw file rather than comparing me to “South Asian”.”
Rohan D. · London
“Upload was two clicks, payment was ordinary, report arrived next morning. The access-code step felt right for DNA.”
Kavya R. · Hyderabad
“My file processed but no model survived — and the report explained exactly why. Not the answer I wanted, but a real one.”
Vikram T. · Chandigarh
“Standard errors on the weights. Competing model listed. This is how ancestry should have been reported all along.”
Ananya B. · Kolkata
“Sent my AncestryDNA zip straight in. No account, no subscription, no follow-up marketing. Just the report.”
Sameer J. · Dubai
Upload your file US$19.99 · one report · by email within 24–48 hours