.bed / .bim / .fam files, for 1kg, converted from the raw data on the PLINK website

# Download the hg38 genome reference files from the PLINK website

RUN wget -L https://www.dropbox.com/s/j72j6uciq5zuzii/all_hg38.pgen.zst

RUN wget -L https://www.dropbox.com/scl/fi/fn0bcm5oseyuawxfvkcpb/all_hg38_rs.pvar.zst?rlkey=przncwb78rhz4g4ukovocdxaz -O all_hg38.pvar.zst

RUN wget -L https://www.dropbox.com/scl/fi/u5udzzaibgyvxzfnjcvjc/hg38_corrected.psam?rlkey=oecjnk4vmbhc8b1p202l0ih4x -O all_hg38.psam

# Download the hg38 related samples file from the PLINK website

RUN wget -L https://www.dropbox.com/s/4zhmxpk5oclfplp/deg2_hg38.king.cutoff.out.id

# Decompress the genome reference files

RUN /plink-ng-master/2.0/bin/plink2 --zst-decompress all_hg38.pgen.zst all_hg38.pgen

RUN rm all_hg38.pgen.zst

RUN /plink-ng-master/2.0/bin/plink2 --pfile all_hg38 vzs --allow-extra-chr --chr 1-22 --max-alleles 2 --remove deg2_hg38.king.cutoff.out.id --memory 6000 --make-bed --out 1kg_hg38

# Replace rsIDs with chr:pos:ref:alt

RUN awk 'BEGIN{OFS="\t"} {print $1,$1":"$4":"$6":"$5,$4,$6,$5}' 1kg_hg38.bim > 1kg_hg38_clean.bim

RUN mv 1kg_hg38_clean.bim 1kg_hg38.bim

# Apply PLINK filtering (mAF > 0.1%, HWE p-value <1e-12, keep SNPs only)

/plink-ng-master/2.0/bin/plink2 --bfile 1kg_hg38 --maf 0.001 --hwe 1e-12 --snps-only --make-bed --out 1kg_hg38_filtered --memory 6000

# Compress output files

gzip 1kg_hg38_filtered.* -v --force