1000 Genomes Project hg38, phased, biallelic, normalized

Pre-processed genotypes from the 1000 Genomes Project. Starting from the per-chromosome phased VCFs found here: left-align & normalize against hg38, atomize, and split multiallelics using bcftools concatenate all chromosomes together convert to PLINK 2 format, compressing the PVAR file with zstd to be compatible with PLINK 2 compress the PGEN file with xz PLINK -> genoray SVAR -> tar archive -> xz compress Note that the SVAR file is ~90 GB decompressed and must be decompressed for use. 32 GB of RAM is recommended for decompression to avoid an out-of-memory error.