Back

Genomic privacy risks in GWAS summary statistics

Lan, A.; Pawitan, Y.; Shen, X.

2025-09-12 genetic and genomic medicine
10.1101/2025.09.09.25335252 medRxiv
Show abstract

The rapid advancement in sequencing technologies has exponentially increased the availability of genomic data, heightening concerns about data privacy. Despite the perceived safety of publicly accessible genome-wide association study (GWAS) summary statistics, we demonstrate that their combination with less sensitive high-dimensional phenotype data can lead to significant leakage of confidential genomic information. By transforming a linear regression model into linear programming constraints, we scrutinize the potential for genomic data recovery using GWAS summary statistics. We found that an effective phenotype-to-sample size ratio above 0.85 could enable full genotype recovery, and that above 0.16 was sufficient to enable individual identification. Certain non-European populations are especially vulnerable. The results stress the urgent need for stronger privacy protections in genomic research while maintaining data utility.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.