GREML-LDMS by chromosome

[PoojaMiddha]

Hi,
I am considering running GREML-LDMS for a large number of samples. My plan is to conduct Steps 1, 2, and 3 separately for each chromosome based on minor allele frequency (MAF) and LD bins. Then, I will use --mgrm for the REML part. I am worried that this method might introduce bias to the heritability estimates.

Thanks.