search

qlu-lab / PUMAS

Fine-tuning polygenic risk score models using GWAS summary statistics
MIT License
41 stars 11 forks source link

SNP weights files #11

Closed lgarvert closed 1 year ago

lgarvert commented 1 year ago

Hi,

I have a question regarding the number von SNP weights files needed to evaluate PRS performance. From the readme I got the impression that only a single file is required (example given was Height.lassosum.txt) but I was wondering how I am supposed to combine the information from all k Iterations in a single file. Looking into the code of PUMAS.evaluation.R (line 70) I saw that it seems to read in k files called (analogous to the example above) Height.lassosum.ite1.txt to Height.lassosum.itek.txt.

Does that mean that I need to generate a separate SNP weights file for each Iteration?

Thanks Linda

zijie02 commented 1 year ago

Hi Linda,

Yes you'll need k SNP weights files to obtain k-fold MCCV results from PUMAS.

Best, Zijie