In order to assess the reliability of imputation with an emphasis on the less common SNPs and an interest in evaluating data imputed from different platforms, we introduce a new statistic, the imputation quality score (IQS). Partly motivated by Cohen's statistic Kappa to quantify rater agreement[15], IQS takes chance agreement into account and thus controls for allele frequencies. In this paper, we introduce IQS, demonstrate its value in situations of low minor allele frequencies, and demonstrate how it can be used to improve the type I error rate when cases and controls are genotyped on different platforms.
