Hi all,

As GWAS provided lots of SNPs that are associated with phenotypes, researchers always applied meta-analysis with much more samples.

Generally, researchers applied regression analysis on the associated variants, and WGRS risk profiles will be used to define the relative risks regarding to baseline.

When I simulated 10,000 random samples based on the MAFs, my question is how to stratify samples into different risk level. If I cut the profile into three categories, I can define 40-60% quantiles as normal, while the other two ones are subjected to low and high risk.

My question is: 1. Are there really individuals with low risk according to risk variants, or "these people really carry protective alleles"? 2. If I need to cut them into 5 categories, is there any existed algorithm that I can apply to do such?

Reference: "Developing and evaluating polygenic risk prediction models for stratified disease prevention"

Thanks, Junfeng