MCPCA_PopGen

MCPCA_PopGen infers population structure from low-depth sequencing genotype dosages using optimized nonlinear transformations while accounting for genotype-calling uncertainty.

Key Features:

• Nonlinear Dimension Reduction: Optimizes nonlinear transformations of genotype dosages to maximize the Ky Fan norm of the covariance matrix, capturing underlying population structure.
• Incorporation of Uncertainty: Integrates genotype-calling uncertainty between heterozygotes and common homozygotes at loci with rare alleles.
• Statistical Power Maintenance: Maintains robust statistical power in low-depth sequencing datasets by optimizing data transformations to reveal hidden population structures.

Scientific Applications:

• Population genetics with low-depth sequencing: Inferring population structure in studies that leverage large sample sizes from low-depth sequencing.
• Single-chromosome analyses: Recovering hidden population structure from single-chromosome data, as demonstrated on samples from two indigenous Siberian populations.
• Human ancestry and migration studies: Investigating patterns of human ancestry and migration through inferred population structure.

Methodology:

Optimizes nonlinear transformations of genotype dosages to maximize the Ky Fan norm of the covariance matrix and incorporates genotype-calling uncertainty between heterozygotes and common homozygotes at loci with rare alleles.