Secure genome-wide association analysis using multiparty computation.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genome data is often restricted, limiting its use in large-scale studies like genome-wide association studies (GWAS).
• Privacy concerns hinder data sharing and public contribution of genomic information.
• Existing secure analysis methods do not scale to computationally intensive tasks such as GWAS.

Purpose of the Study:

• To develop a scalable protocol for secure genome-wide analysis.
• To enable privacy-preserving quality control and population stratification correction in large cohorts.
• To facilitate the use of restricted genomic data and promote secure genome crowdsourcing.

Main Methods:

• A novel protocol for large-scale genome-wide analysis was developed.
• The protocol was tested for quality control and population stratification correction in cohorts of 9K, 13K, and 23K individuals.
• The scalability of the protocol was assessed for potential application to a million individuals.

Main Results:

• The protocol successfully performed quality control and population stratification correction while maintaining genotype and phenotype confidentiality.
• The method demonstrated feasibility for scaling to large cohorts, including up to 23K individuals in current tests.
• The approach shows potential to scale to a million individuals for genome-wide analyses.

Conclusions:

• The developed protocol enables secure, large-scale genome-wide analysis, addressing privacy concerns.
• This method can unlock access to restricted genomic datasets for research.
• The protocol supports secure genome crowdsourcing, encouraging wider participation in genetic studies without compromising privacy.