search

choishingwan / PRS-Tutorial

A tutorial on how to run basic polygenic risk score analysis
MIT License
68 stars 104 forks source link

Confusion about pruning and clumping #53

Closed ptranvan closed 7 months ago

ptranvan commented 11 months ago

In the PRS Plink section, there is a clumping step, but I thought we already had a set of uncorrelated SNPs because in the QC of target data section, we had performed a pruning step.

So, I am confused about why there is a clumping step after pruning. I thought it was either one or the other, but can we actually perform both steps together?

choishingwan commented 7 months ago

prunning was done for QC, mainly for sex check and heterozygous analyses. We would not thin down the genotype file based on the pruned SNPs for subsequent analyses.

For the actual PRS calculation, the full genotype were used, and clumping was then done to extract the most significant SNPs within the region.