Identifying disease-causing mutations with privacy protection.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genome data is crucial for disease diagnosis and treatment.
• Sharing sensitive genome data raises significant privacy and security concerns due to re-identification risks.
• Access to large-scale genomic datasets is essential for interpreting patient genomes and identifying disease-gene relationships.

Purpose of the Study:

• To develop a method for identifying disease-associated variants and genes while preserving patient privacy.
• To enable secure analysis of genomic data for medical diagnosis without compromising individual confidentiality.

Main Methods:

• Utilizes secure multi-party computation (SMC) for privacy-preserving analysis.
• Identifies disease-causing mutations under various inheritance models (recessive, dominant, compound heterozygous).
• Compares re-implemented privacy-preserving methods (MAX, SETDIFF, INTERSECTION) for efficiency.

Main Results:

• The prototype analyzes thousands of genomic datasets in milliseconds with logarithmic runtime scaling.
• Achieves significant speed improvements: MAX (2.5x), SETDIFF (1122x), and INTERSECTION (341x) faster than the state-of-the-art.
• Demonstrates the first privacy-preserving genomic analysis based on inheritance models for mutation detection.

Conclusions:

• The proposed SMC approach effectively identifies disease-associated variants and genes while safeguarding patient privacy.
• The method significantly enhances the speed and efficiency of privacy-preserving genomic data analysis.
• Enables broader and more secure utilization of genomic data in clinical settings.