Public platform with 39,472 exome control samples enables association studies without genotype sharing.

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genomics

Background:

• Recruiting matched control samples for genetic association studies is often challenging and resource-intensive.
• Existing genetic data sharing faces significant hurdles due to strict privacy regulations for human genetic information.

Purpose of the Study:

• To develop a method for selecting optimal control samples from external datasets without requiring direct genotype sharing.
• To enhance the power of genetic association studies by leveraging readily available control data.

Main Methods:

• Utilized singular value decomposition and a subsampling algorithm to identify best-matching controls.
• Developed an online library of 39,472 exome sequencing controls accessible at http://dnascore.net.
• Implemented a system for selecting control sets with prespecified matching accuracy.

Main Results:

• Successfully demonstrated a method for selecting appropriate controls from a large external dataset.
• The developed method ensures compliance with personal data protection regulations.
• Enabled association analyses for case cohorts that previously lacked adequate control subjects.

Conclusions:

• The developed method significantly improves the feasibility and power of genetic association studies.
• Provides a valuable resource for researchers needing control cohorts, especially for rare variant analysis.
• Facilitates well-calibrated genetic association analyses through accurate control sample selection.