PSMix

PSMix infers population structure and individual admixture from multilocus genomic marker data using a maximum likelihood framework implemented via the expectation-maximization (EM) algorithm in R.

Key Features:

• Efficient computation: Uses an expectation-maximization (EM) maximum likelihood approach that is computationally more efficient than Bayesian MCMC-based methods such as STRUCTURE.
• Accurate inference: Produces admixture and population stratification estimates with accuracy reported comparable to established Bayesian software.
• R implementation: Provided as an R package for analysis of multilocus genotypes.

Scientific Applications:

• Association mapping: Accounts for population stratification and individual admixture to reduce false positives in genetic association studies.
• Evolutionary studies: Characterizes population structure and admixture patterns to inform analyses of species evolution and migration.

Methodology:

Applies maximum likelihood estimation via the expectation-maximization (EM) algorithm to multilocus genotype data and contrasts with Bayesian MCMC approaches that can exhibit convergence issues.