Next Generation of Genotype Imputation Methods

• Activity Number: 92
• Type: Invited
• Date/Time: Sunday, August 4, 2013 : 8:30 PM to 10:30 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Sayantan Das*+ and Goncalo R. Abecasis
• Companies: University of Michigan and University of Michigan
• Keywords: Imputation ; Statistical Genetics ; Hidden Markov Model ; Haplotypes ; Methods

Abstract

Genotype imputation is a key step in the analysis of human genetic studies as it facilitates in increasing power of gene mapping, enables combination of results across different studies and accelerates fine-mapping efforts. Imputation works by finding the shared haplotype segments between the sample individuals typed on a commercial array and a reference panel of more densely typed individuals (e.g. The International HapMap Project). Advanced technologies in high-throughput sequencing have resulted in the rapid increase in size of these publicly available data sets. Using these as reference panels would soon impose a computational burden. We introduce a strategy called 'state space reduction' which reduces the description of short genomic regions to the number of distinct haplotypes rather than the total number of haplotypes. The existing formulas have been refined to work with only these distinct haplotypes in a series of short genomic regions which covers the whole genome. The proposed algorithm maintains the accuracy of the current methods while reducing the computational cost. We also formulate a recursive algorithm to find the optimal allocation of such genomic regions.