A haplotype map derived from whole genome low-coverage sequencing of over 25,000 individuals. J. Marchini on behalf of the Haplotype Consortium Department of Statistics, University of Oxford, United Kingdom.
Several whole-genome low-coverage sequencing studies of population and disease cohorts are currently underway or nearing completion. Each of these studies will produce haplotype sets that can be used for downstream imputation into other GWAS cohorts. The goal of the Haplotype Consortium is to combine the data across 18 studies totaling more than 25,000 predominantly European samples, resulting in a reference set of haplotypes that will act as a central resource for imputation and population genetics studies. The large sample size of the combined study will improve the accuracy of the genotype calling and phasing in each individual study. This will limit the need for groups to choose between reference sets or carry out imputation from multiple reference sets and, importantly, should improve imputation of rare variants. There are several challenges in achieving the goals of the Haplotype Consortium. For example, the haplotype sets produced by each individual study are not easily combined due to differences between sequence analysis protocols and resulting lists of polymorphic sites identified in each study. We have investigated strategies for combining haplotype sets using phasing and imputation methods, by combining data at the haplotype or genotype level. In addition, we have investigated approaches that use genotype likelihoods at a union set of polymorphic sites across studies. I will present the results of these comparisons together with our plans for how this resource can be used by the community for imputation based analysis, as well as our future plans for extending the size of this resource to include other ethnicities.
You may contact the first author (during and after the meeting) at