Genotype-Privacy-PRS-Benchmark

We ran polygenic risk score (PRS) calculations using 1000 Genome Project's Omni Microarray genotype data on a population of 2318 individuals and show the ancestry and PRS results here. The calculation and related data preprocessing are done by the Michigan Imputation Server.

Data Sources

1. 1000 Genome Genotype Data
   1. To download, run

      curl -O ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/supporting/hd_genotype_chip/ALL.chip.omni_broad_sanger_combined.20140818.snps.genotypes.vcf.gz
      

2. 1000 Genome Phase 3 v5 Reference Panel for running imputation & PRS calculations

Data Preparation

After downloading the Omni genotype data hosted by 1000 Genome, run the following to dissect the whole genome VCF file into ones arranged by chromosomes:

for chr in {1..22}; do
    bcftools view -r ${chr} -Oz -o chr${chr}.vcf.gz \
    ALL.chip.omni_broad_sanger_combined.20140818.snps.genotypes.vcf.gz
done

Running PRS Calculation

Upload the prepared per-chromosome vcf.gz files to the Michigan Imputation Server and select configurations accordingly (see below).

• Genotype data is prepared according to genome build GRCh37/hg19.
• Use 1000G Phase 3 v5 as reference panel for the job.
• Feel free to adjust rsq filter and trait category for polygenic scoring.
  • For the benchmark results shown in this repo, trait category is Cancer.