BIOSTAT 666 Final Project: Comparing Polygenic Risk Scores Calculated from GWAS Summary Statistics

Authors: Gabrielle Dotson, Mengtong Hu, Jeremy Kaplan, and Soumik Purkayastha

Setup

Conda Environment

All of the library requirements for this project can be installed in a conda environment. From a command line you can build the environment as

conda env create -f environment.yml

Once the environment builds, you can activate the environment by running:

conda activate b666prj

Data

Reference genotypes were downloaded from the 1000 Genomes Project FTP Server Download files to input/vcf

GWAS summary statistics were downloaded from the UNC Psychiatric Genomics Consortium. Download and unzip files into input/gwas

Running the workflow

Currently the workflow will process VCFs and generate clumped SNPs on the merged denotypes. You can run the entire workflow locally by running:

snakemake

On a cluster, where you can run the jobs in parallel you can run it as:

snakemake --cluster "sbatch -A {my_account} " -j {n_chromosomes}

where my_account is the slurm account you want to bill, and n_chromosomes is the number of chromosomes you want to run in parallel(probably 22)

This will create a number of outputs in output/ containing information about the clumped SNPs.