Hi! I'm new to GWAS and have a question regarding LMM-based GWAS software, specifically REGENIE (https://rgcgithub.github.io/regenie/).

Can LMM GWAS handle sample relatedness (i.e., population structure) when there are very unequal sample sizes in terms of anestry? Could this lead to problems or are there additional steps/checks that should be taken in to account?

For context, my sample includes 15,873 participants. Split by ancestry, this is equivalent to: 76 AFR, 18 AMR, 50 EAS, 15,616 EUR, and 113 SAS.

Thanks for your help :)