Estimate the GRM- use genotyped or imputed data?

new

« Prev
1
Next »

irene
New Member

Posts: 5

Estimate the GRM- use genotyped or imputed data? Sept 16, 2013 16:38:16 GMT

Quote

Post by irene on Sept 16, 2013 16:38:16 GMT

Dear all,

I am a new user of GCTA and I would like (justified) reasons for some options that are available.

Should we use only genotyped or can we also use imputed data (with filtering on R2) to estimate the GRM?
What are the risks/ benefits?

Many thanks in advance.

apple
New Member

Posts: 1

Estimate the GRM- use genotyped or imputed data? Sept 17, 2013 12:00:15 GMT

Quote

Post by apple on Sept 17, 2013 12:00:15 GMT

I would use common SNPs on HapMap3 regardless of whether the data are imputed to HapMap2, HapMap3 or 1000G. The benefit is that the HapMap2 SNPs should cover the genome a bit better than genotyped SNPs and you don't need to worry about the issue of uneven coverage as in HapMap2 and 1000G.