AncestryDNA raw data analysis
Your AncestryDNA file holds more than ancestry
AncestryDNA was built to trace where your family came from. The raw genotype file it generated also sits on clinically relevant positions a health service would read, and Ancestry never interprets those for you. Re-analyse the file you already own, in the EU, and see what is in it.
Re-analyse your AncestryDNA file · €19 See a sample report first
€19 one-time · upload the .zip Ancestry's download link gives you · processed in the EU, never sold.
The variants Ancestry doesn't report
AncestryDNA's product is ancestry composition and DNA matches. Its genotyping chip, though, reads hundreds of thousands of positions, and a subset of those are ones Ancestry never interprets for you: APOE, Lp(a), HFE and APOL1, alongside the CPIC pharmacogenomic star alleles. Nothing new is sequenced. The information is already in the file you downloaded, and re-analysis is what turns it into an attributed, plain-language report.
How to download your AncestryDNA raw data
- Sign in at Ancestry and go to your DNA settings (via your account/DNA menu).
- Choose Download DNA Data (sometimes shown under the settings for your specific test).
- Confirm your password; Ancestry emails you a secure link.
- Follow the link to download the raw file (a zipped
.txt). That archive is what we re-analyse, so keep it somewhere private.
Where the Ancestry array differs from 23andMe's
AncestryDNA uses a different genotyping array, and its export is shaped differently too: two allele columns per marker rather than a single genotype string, on build 37, with no build line in the header to say so. The set of positions read differs as well, so some pharmacogenomic variants are covered and others aren't. The recessive-carrier panel is not a chip product at all, on any array.
We read the positions your specific file contains and attribute every finding to ClinVar, ACMG or CPIC. Positions your chip didn't read are shown as no-calls, never imputed over. Structural variation stays out of reach either way: CYP2D6 duplications aren't resolvable from array data, so where a metaboliser status would be uncertain we mark it indeterminate rather than guess.
A whole-genome re-analysis resolves much of what an array can't. For most people the AncestryDNA file is a genuinely useful, honest starting point.
What we keep, and for how long
Your upload is processed and stored in Finland and never leaves the EU, under GDPR Article 9 special-category protection. Ancestry's download link already put the archive on your own disk; our copy is deleted after analysis by default, and no genome is held here that a court could sell. Security & your data. If Ancestry's own data handling is why you're here, the 23andMe bankruptcy explainer sets out why "who holds your genome" is the question that decides the rest.
The cheaper option, and what it costs later
Promethease and Genetic Genie read the same public databases we do. The difference is time. They hand you a report frozen on the day you run it, read from the few hundred thousand positions a genotyping chip types. We read your whole genome where you have it, attribute every finding to the guideline it came from, and can re-read your file as those guidelines change. Your report is a one-off purchase and stays yours. Keeping it current as the guidelines move is the optional re-analysis subscription, and you can start or stop that whenever you like.
Named services are described from their public product pages; features change and we don't speak for them. Three years later, that report still says exactly what it said on the day you ran it.
Ancestry never interprets these positions. One re-analysis of your file is €19.
Other file types: 23andMe raw data · MyHeritage raw data · Choosing a service that never holds your genome.