23andMe Publication Dataset Access Program
23andMe
Summary statistics from one of the largest consumer DNA cohorts, but only through a signed contract.
23andMe shares full genome-wide association summary statistics from its published studies, built from customers who consented to research. Academic and nonprofit researchers get them free after a request form, a Data Transfer Agreement, and a Statement of Work signed by their institution; commercial institutions have a separate route. Individual-level genotypes are not offered.
Access
- Cost
- By applicationFree for academic and nonprofit institutions after a reviewed request, DTA, and SOW (about four weeks for signing); only named researchers may use the data; publication limited to statistics for up to 10,000 SNPs.
- License
- Proprietary23andMe Data Transfer Agreement
- Sign-up
- Application required
- How you get it
- Through the provider's siteTSV
What's in it
- One row is
- One SNP's association statistics for one published phenotype
- Covers
- Mainly U.S. customers who consented to research
- Years
- Studies published since about 2010
- Updated
- Irregular
- Size
- Unknown; one summary statistics file per published GWAS
Quality
Very large samples make for well-powered GWAS. Phenotypes are self-reported survey answers and the customer base is wealthier and more European-ancestry than the population.
- Self-reported phenotypes
- Customer base not representative
- Contractual limits on what can be published
Connects to
- GWAS Catalog: Published 23andMe associations are curated there
- gnomAD (Genome Aggregation Database): Check allele frequencies of associated variants
Questions it could answer
- How do 23andMe effect sizes compare with biobank-based GWAS of the same trait?
- Which self-reported traits show the strongest genetic signals?