A major goal of performing imputation in Scenario B (and extensions thereof) is to simultaneously use all available reference data in an integrated modeling framework. As such, it is also important to assess the quality of imputation at SNPs in U1 (i.e., HapMap-only SNPs) in this context. To do so, we created a modified version of the full Scenario B dataset with observed Illumina genotypes in the study sample. We masked every 25th Illumina SNP in both the study sample and the diploid reference panel, then ran BEAGLE and IMPUTE v2 as before. We repeated these steps for each of the 24 other possible sets of masked SNPs (i.e., after shifting the starting index), so that every Illumina SNP was masked and imputed exactly once.
