bigsnpr

bigsnpr provides memory-efficient analysis of large-scale SNP array and genotype data within R for population genetics and genome-wide association studies.

Key Features:

• Memory-Mapping Technology: Uses memory-mapping to access SNP and genotype matrices on disk rather than loading them entirely into RAM.
• Integration with External Tools: Integrates with external genomic software via transparent system calls or by implementing updated methods in R.
• Principal Component Analysis (PCA): Implements fast and accurate PCA computations for population-structure inference and confounding correction.
• Genome-Wide Association Studies (GWAS): Performs GWAS to identify genetic variants associated with traits and diseases.
• Linkage Disequilibrium (LD) Pruning: Provides functions to prune SNPs in linkage disequilibrium to retain independent markers.
• Polygenic Risk Scores (PRS): Provides algorithms to compute polygenic risk scores from millions of SNPs.
• Scalability: Demonstrated analysis of a simulated dataset comprising 500,000 individuals and one million markers on a single desktop.

Scientific Applications:

• Large-Scale GWAS and Case-Control Studies: Enables genome-wide association analyses, including case-control studies such as analyses of celiac disease.
• Population Structure Analysis: Facilitates PCA-based analyses to characterize population structure and correct for stratification.
• Polygenic Risk Prediction: Supports computation of PRS for disease susceptibility and trait prediction from high-density SNP data.

Methodology:

Uses memory-mapping for on-disk data access, performs transparent system calls to external software or implements methods in R, and provides fast routines for PCA, GWAS, LD pruning, and PRS computation.