What to Do With 23andMe Raw Data in 2026: The Complete Analysis Guide
Download your 23andMe file and upload it for disease risk, carrier screening, pharmacogenomics, and ACMG findings you won’t see in the app. Step-by-step guide to getting the most from your DNA.
Category: Diagnostics · By GenomeNarrator Team · · 8 min read
When 23andMe sends you a health report, you're seeing a small fraction of what your data actually contains. The raw data file, that plain-text download buried in your account settings, holds around 600,000 to 700,000 genotyped positions. 23andMe reports on a few hundred of them. The rest sits unused unless you take it somewhere else.
What's inside your raw data file
Your 23andMe raw data is a tab-separated text file. Each row is a single SNP (single nucleotide polymorphism), a position in your genome where people commonly differ. The file contains your rsid (a universal variant identifier like rs429358), the chromosome, the position, and your genotype (the two letters you inherited, one from each parent).
This file doesn't contain your full genome; it's a curated panel of SNPs chosen because they tag important regions. But 600K+ variants is enough to calculate meaningful risk scores across hundreds of conditions.
Disease risk analysis
The most valuable use of raw data is calculating disease risk across conditions that 23andMe doesn't cover. 23andMe's FDA-cleared health reports are limited to a handful of conditions: roughly 10 to 15 depending on your plan. But your raw data contains variants relevant to hundreds more.
Tools like GenomeNarrator scan your raw data against ClinVar (the NIH clinical variant database), the GWAS Catalog (thousands of genome-wide association studies), and curated pharmacogenomic guidelines. The result is a scored risk assessment across 1,000+ conditions: from coronary artery disease and Type 2 diabetes to rarer conditions like alpha-1 antitrypsin deficiency.
Carrier screening
Your raw data can reveal whether you carry recessive variants for conditions like cystic fibrosis, sickle cell disease, Tay-Sachs, or hereditary hearing loss. Being a carrier doesn't affect your own health, but it matters if your partner carries the same condition: your children could be affected.
23andMe covers some carrier conditions, but the raw data contains variants for many more. A thorough carrier analysis can flag conditions you'd otherwise only discover through expensive clinical panel testing.
Pharmacogenomics: how you respond to medications
Some of the most immediately actionable information in your raw data is pharmacogenomic. Variants in genes like CYP2D6, CYP2C19, DPYD, and SLCO1B1 affect how your body metabolizes specific drugs.
For example, roughly 2% of people are CYP2D6 poor metabolizers: codeine provides them zero pain relief because they can't convert it to morphine. About 30% of people of East Asian descent carry CYP2C19 variants that change how they process the blood thinner clopidogrel (Plavix). These aren't theoretical risks. They affect prescribing decisions happening today.
GenomeNarrator flags these drug-gene interactions using CPIC (Clinical Pharmacogenetics Implementation Consortium) Level A guidelines, the same evidence base used by major hospital pharmacogenomics programs.
ACMG secondary findings
The American College of Medical Genetics maintains a list of 84 genes where incidental findings are considered medically actionable, meaning there's something a doctor can do about it right now. These include BRCA1/2 (breast and ovarian cancer risk), LDLR (familial hypercholesterolemia), and several cardiac arrhythmia genes.
Your 23andMe raw data covers some of the SNPs in these genes. A secondary findings screen checks whether you carry known pathogenic variants in any of these 84 genes and flags them for clinical follow-up.
Trait analysis
Beyond health, your raw data encodes information about physical and behavioral traits: eye color, caffeine metabolism, sleep chronotype, bitter taste perception, and more. These are lower stakes but often the most immediately engaging part of a report.
What you should not expect
Raw data from SNP arrays has real limitations. It doesn't cover structural variants, copy number variations, or most rare mutations. If a variant isn't on the 23andMe chip, it won't appear in your file. This means a "clear" result doesn't guarantee you don't carry a pathogenic variant; it means the array didn't test for it.
Additionally, polygenic risk scores derived from SNP array data are most accurate for people of European descent because the underlying GWAS studies disproportionately sampled European populations. Scores for other ancestries are improving but remain less precise.
How to download your raw data
Log into your 23andMe account, go to Settings, scroll to "23andMe Data," and click "Download Raw Data." You'll receive a zip file containing a single text file. This file works with GenomeNarrator, Promethease, and other third-party analysis tools.
The bottom line
Your 23andMe raw data is worth far more than the reports 23andMe gives you. A single file upload to a tool like GenomeNarrator produces a comprehensive analysis (disease risk across 1,000+ conditions, carrier screening, pharmacogenomic drug interactions, ACMG secondary findings, and trait analysis) in under a minute. The data is already yours. The question is whether you're using all of it.
More from GenomeNarrator: all genomics articles, how the analysis works, or GenomeNarrator home.