Identifying the phenotypic effect of rare variants by including between-pedigree coancestry in variance components linkage analysis

Abstract

The contribution of rare variants to disease burden has become an important focus in genetic epidemiology. These effects are difficult to detect in population-based datasets, and as a result, interest in family-based study designs has resurfaced. Linkage analysis tools will need to be updated to accommodate the scale of data generated by modern genotyping and sequencing technologies.\n\nIn conventional linkage analysis individuals in different pedigrees are assumed to be independent of each other. However, cryptic relatedness is often present in populations and haplotypes that harbor rare variants may be shared between pedigrees as well as within them.\n\nWith millions of polymorphisms, Identity-by-descent (IBD) states across the genome can now be inferred without use of pedigree information. This is done by identifying long runs of identical-by-state genotypes which are unlikely to arise without IBD. Previously, IBD had to be estimated in pedigrees from recombination events in a sparse set of markers.\n\nWe present a method for variance-components linkage that can incorporate large number of markers and allows for between-pedigree relatedness. We replace the IBD matrix generated from pedigree-based analysis with one generated from a genotype-based method. All pedigrees in a dataset are considered jointly, allowing between-pedigree IBD to be included in the model.\n\nIn simulated data, we show that power is increased in the scenario when there is a haplotype shared IBD between members of different pedigrees. If there is no between-pedigree IBD, the analysis reduces to conventional variance-components analysis. By determining IBD states by long runs of dense IBS genotypes, linkage signals can be determined from their physical position, allowing more precise localization.