Dear all, when I project the the imputed 992 ANMerge subjects along with 1000 genome individuals. I found a deviation of imputed subjects with EUR population, which indicates some potential bias in the imputed genotyping data. I did not find any imputation quality file either. Does anyone has the same problem? I attached a image in dropbox link (https://www.dropbox.com/s/it3ns26wmeu4xg3/ANMerge992_1000G.png?dl=0).