or differing types of genetic information (i.e., differing SNP selection strategies used for Illumina and Affymetrix), one would expect differing imputation accuracy results from the different arrays. Combining imputation across arrays with differing inputs seems likely to generate systematically differential imputation error among individuals contributing to the composite data set and thus the observed biases, including greater bias among SNPs with lower MAF, for which genetic information on which to base prediction of imputed SNPs is less.
