Patterns of IBD sharing inferred from whole genome sequences of 962 European Americans. C. V. Van Hout1, F. Yu2, X. Liu3, E. Boerwinkle2,3, A. G. Clark1 1) Molecular Biology and Genetics, Cornell University; 2) Human Genome Sequencing Center, Baylor College of Medicine; 3) Human Genetics Center, University of Texas Health Science Center at Houston.
Past demography of a population is expected to leave a signature in the distribution of numbers and lengths of segments of the genome that are shared identity by descent (IBD) between individuals. Recently, methodological advances and the availability of whole genome sequencing data for large numbers of individuals within a single population have enabled high resolution inference of IBD block sharing, allowing investigation of recent human genealogic history. We identified pairwise-shared IBD blocks with the Beagle software (Browning & Browning, Genetics, 2013; Gusev et al., Genome Res., 2009) in whole genome sequence in a sample of 962 European Americans from the ARIC study. Assuming unilineal relationships, we estimated empirical pairwise kinship coefficients as a function of the genetic lengths of shared blocks. The distribution of kinship coefficients for 462,241 pairs of putatively unrelated study participants had an inter-quartile range of 0.00006-0.0003 and mean 0.0002 (between 5th and 6th cousins). This implies that in this sample, on average 92 pairs of participants are expected to share any given locus IBD. Recent origins of common ancestry would be expected to yield long blocks of IBD sharing, and as the time back to a common ancester grows, we expect the length of the IBD block to shrink at the same time that sequence differences will accumulate by mutation. We summarized the distribution of shared block lengths and the number of nucleotide differences in each block. Consistent with established models of human population history, we observed a block length distribution heavily skewed toward shorter segments. In addition, we observed an increase in the mean and variance of the number of nucleotide differences per kilobase in shorter shared IBD blocks, consistent with their increased age. The mutation accumulation within IBD blocks and block length reduction by recombination provide types of molecular clocks whose consistency under specified demographic models was tested. We conclude that IBD block inference from whole-genome sequence data provides and exceptionally rich platform for analysis of recent genealogic kinship and demography.
You may contact the first author (during and after the meeting) at