A Snakemake pipeline for calculating polygenic risk scores from the PGS-Catalog on large sets of genomic data in plink2 binary format. This pipeline was developed for use with data from the UK-Biobank. First, risk score variant weights are downloaded from PGS-Catalog, then lifted to the correct reference genome and formated such that plink2 can compute per sample risk scores in the final step.

For use within the lab you can activate the conda environment prs:

conda activate prs

or use the environment.yaml to create a new conda environment with the necessary dependencies.

Given the polygenic score ID PGS001954 for HDL-Cholesterol:

snakemake /some_dir/PGS001954.sscore

computes the polygenic risk scores for PGS001954 on the plink2 genomic data specified in the config.yaml.

The column SCORE1_AVG in /some_dir/PGS001954.sscore then contains the risk scores.