PGS Catalog file does not have rsID
1
0
Entering edit mode
4 months ago
Jenish ▴ 20

Hi all, I am trying to use a PGS Catalog file (PGS000013) to calculate PRS scores. The PGS file that I am using has ~6 million variants and does not have rsID. The issue is that using a python library or an api to map each of the 6 million variants to a rsID is really really slow, which is why I am not able to calculate the PRS Score using that file with PRSice. Is there a faster way to map chromosome positions to a rsID or are there any other repositories similar to pgscatalog.org?

Catalog CAD PGS PRSice • 347 views
0
Entering edit mode
4 months ago
Sam ★ 3.8k

Hi, with later version of PRSice (e.g. 2.3.5), you can use the --chr-id c:l option to do the mapping automatically. If not, you can maybe modify the bim file and your PGS catalog files to have ID of chr:loc.

0
Entering edit mode

Thanks. Should there be any difference in how this would work vs when using SNP ids? The c:l to snp id mapping should be 1 to 1 right?

1
Entering edit mode

The actual implementation of chr-id so far is rather iffy, in that if I remember correctly, the extraction and exclusion doesn't work as expected, and the chr-id will not work if you have a SNP ID for your summary statistic file. So in short, it works in place of the normal SNP id matching. Am working to make it more intuitive for the next release.