pg_gpu

GPU-accelerated population genetics analysis using the pg_gpu library (https://github.com/kr-colab/pg_gpu). Use this skill whenever the user wants to: load a VCF or zarr file for population genetics analysis, calculate diversity statistics (pi, Tajima's D, theta), run selection scans (iHS, XP-EHH, nSL), compute divergence/FST/dxy between populations, calculate windowed statistics across chromosomes, handle missing data, compute the site frequency spectrum, run PCA or dimensionality reduction, plot windowed stats across chromosomes, panel multiple chromosomes, or run any pg_gpu function. Always use this skill when the user references @vcf, population genetics stats, or pg_gpu commands.