Several imputation programs are available, and our results showed that the highest genotype imputation accuracy and quality were achieved in African Americans using IMPUTE2. The IMPUTE2 program is also computationally efficient [17]. Chanda et al. [8] previously reported that IMPUTE and MaCH performed equally well in African Americans and that both programs performed consistently better than BEAGLE, regardless of the 1000 Genomes reference panel used from the August 2010 and June 2011 releases. Our comparisons of the latest IMPUTE version (i.e., IMPUTE2), BEAGLE, MaCH, and MaCH-Admix with reference to more recent 1000 Genomes panels (February 2012 release) showed the same pattern as Chanda et al. [8] when using the concordance metric to evaluate imputation accuracy. However, concordance can overestimate the agreement of imputed and true genotypes when the MAF is low, due to random chance [30]. By taking the observed agreement (i.e., concordance) and subtracting out the chance agreement derived from the SNP’s allele frequencies, IQS is less prone to overestimating the imputation accuracy of low MAF SNPs. Our calculation of the IQS metric demonstrated an advantage of using IMPUTE2
