Privacy-preserving GWAS analysis on federated genomic datasets

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genomic data availability fuels biomedical research, including Genome-Wide Association Studies (GWAS).
• High-quality GWAS necessitate large sample sizes, often exceeding single institutional capacity.
• Federated analysis enables cross-institutional collaboration but faces privacy and confidentiality challenges.

Purpose of the Study:

• To develop a privacy-preserving framework for federated genomic data analysis.
• To facilitate secure Genome-Wide Association Studies (GWAS) across multiple institutions.

Main Methods:

• Utilizing secure multi-party computation (MPC) systems to layer GWAS computations.
• Implementing a framework for minor allele frequency counting and chi-squared (χ2) statistics calculation.
• Employing the Portable Circuit Format (PCF) for efficient prototyping.

Main Results:

• Demonstrated a functional framework for privacy-preserving GWAS computations.
• Experimental results show the potential for efficient and secure cross-institution GWAS.
• Successfully performed key GWAS computations like minor allele frequency and χ2 statistics.

Conclusions:

• The proposed framework effectively addresses privacy concerns in federated genomic analysis.
• Enables collaborative GWAS across institutions while maintaining data confidentiality.
• Shows promise for advancing large-scale genomic research through secure data sharing.