PCAmatchR

PCAmatchR computes weighted Mahalanobis distances from user-supplied principal components to match cases and controls by genomic similarity and reduce confounding from population stratification in GWAS and other genetic association studies.

Key Features:

• Principal Component Analysis Integration: Accepts user-supplied PCA outputs to base matching on existing genetic principal components.
• Weighted Mahalanobis Distance Metric: Converts principal components into a weighted Mahalanobis distance metric, assigning weights to each principal component based on the percentage of genetic variation it explains.
• Enhanced Genomic Similarity: Selects controls with increased genomic similarity to cases to minimize population stratification and reduce inflation in association test statistics in GWAS.
• Performance Validation: Functionality and performance have been demonstrated using data from the 1000 Genomes Project.

Scientific Applications:

• GWAS case-control matching: Improves selection of well-matched controls in genome-wide association studies to reduce ancestry-related confounding.
• Ancestry control in genetic studies: Applies to studies that require precise control of ancestry-related genetic variation by matching on principal components.

Methodology:

PCAmatchR converts principal components into a weighted Mahalanobis distance metric, weights components by percentage of genetic variance explained, and uses that distance to select controls based on genomic similarity to cases.