An Evaluation of the MiDCoP Method for Imputing Allele Frequency in Genome Wide Association Studies

Abstract

A genome wide association studies require genotyping DNA sequence of a large sample of individuals with and without the specific disease of interest. The current technologies of genotyping individual DNA sequence only genotype a limited DNA sequence of each individual in the study. As a result, a large fraction of Single Nucleotide Polymorphisms (SNPs) are not genotyped. Existing imputation methods are based on individual level data, which are often time consuming and costly. A new method, the Minimum Deviation of Conditional Probability (MiDCoP), was recently developed that aims at imputing the allele frequencies of the missing SNPs using the allele frequencies of neighboring SNPs without using the individual level SNP information. This article studies the performance of the MiDCoP approach using association analysis based on the imputed allele frequency by analyzing the GAIN Schizophrenia data. The results indicate that the choice of reference sets has strong impact on the performance. The imputation accuracy improves if the case and control data sets are imputed using a separate but better matched reference set, respectively.