Polygenic Risk Score (PRS) analysis to quantify genetic predisposition for a trait of interest using GWAS summary statistics as base data and target genotype data.
Following a rigorous step-by-step workflow based on the Choishingwan PRS Tutorial - https://choishingwan.github.io/PRS-Tutorial/base/, I ensured data integrity and quality throughout the analysis.
Starting with base data obtained from GWAS summary statistics, I confirmed the effect allele orientation, checked for file integrity with md5sum, and performed stringent quality control filtering, removing variants with low minor allele frequency (MAF < 1%) and poor imputation quality (INFO < 0.8). I removed duplicate and ambiguous SNPs to maintain robustness and avoid bias in the PRS calculation.
After QC, I aligned SNP effects with the target dataset considering strand orientation, excluded problematic variants, and prepared the data for PRS computation. Using software such as PLINK and R, I calculated individual PRS scores by weighting the genetic variants by their effect sizes from the base data.
Finally, I analyzed and interpreted the PRS results by associating the scores with phenotype data, performing statistical tests (e.g., t-tests comparing cases versus controls), and visualizing score distributions. This detailed and reproducible approach ensured that the PRS values reliably capture genetic risk relevant to the phenotype, supporting further genetic and epidemiological studies.
This process highlights key quality control checkpoints and data harmonization steps essential for valid PRS analyses, offering a solid foundation for personalized risk prediction and genetic epidemiology research.