Entering edit mode

8 weeks ago

MHSATO
•
0

Hi.

I am trying to calculate PRS by using PRSice-2. I'd like to know how to handle the model. When I provide same gwas results to PRSice-2's --base option, is the model always same?

I have several thousands samples measured snp array and target quantitative phenotype. I'd like to calculate PRS for all of these samples. Followings procedures are my plan.

- GWAS by using randomly selected 3/4 of all samples
- Construct PRS model by using above result and diagnose of the model by using rest sample.
- Then, calculate PRS for all samples by using GWAS result used in step 1 and p-threshold in step 2.

For this procedure, I need to confirm that PRSice-2 always construct same model when I use same GWAS result which provided by --base option.

Otherwise, is there another way to fit the model predicted by PRSice-2 to new dataset?

Thanks.