High-throughput inference of pairwise coalescence times identifies signals of selection and enriched disease heritability

Abstract

Interest in reconstructing demographic histories has motivated the development of methods to estimate locus-specific pairwise coalescence times from whole-genome sequencing data. Here we introduce a powerful new method, ASMC, that can estimate coalescence times using only SNP array data, and is orders of magnitude faster than previous approaches. We applied ASMC to detect recent positive selection in 113,851 phased British samples from the UK Biobank, and detected 12 genome-wide significant signals, including 6 novel loci. We also applied ASMC to sequencing data from 498 Dutch individuals to detect background selection at deeper time scales. We detected strong heritability enrichment in regions of high background selection in an analysis of 20 independent diseases and complex traits using stratified linkage disequilibrium score regression, conditioned on a broad set of functional annotations (including other background selection annotations). These results underscore the widespread effects of background selection on the genetic architecture of complex traits.

Concerted variation of the 3′ regulatory region of Ig heavy chain and Gm haplotypes across human continental populations

“…This finding leads to reconsider the structural and regulatory elements in the entire region as a unit involved in health and disease (Oxelius, 2008;Oxelius & Pandey, 2013). In agreement with this, the region showed signals of recent positive selection (Palamara, Terhorst, Song, & Price, 2018). In this context, adaptation to the local environments (Hancock et al, 2011;Iskow, Gokcumen, & Lee, 2012;Oxelius, 2000) may have been a major drive for divergence.…”

“…As far as recent selection is concerned, Palamara et al () analyzed the UK Biobank data and identified an excess of recent coalescence events in the immunoglobulin complex, which is attributable to positive selection during the past 20,000 years. While this signal is centered telomerically to IGHG3 , the resolution of the method is of the order of megabases and may imply that the genomic region here considered is also involved.…”

“…It has long been realized that loci displaying extremely high population differentiation might be under positive selection (Cavalli-Sforza et al, 1994;Lewontin & Krakauer, 1973). We noticed that the extreme F ST values reported for the IGHG1-IGHG3 system worldwide implicitly favor directional selection over simple genetic drift (Biswas & Akey, 2006) As far as recent selection is concerned, Palamara et al (2018) analyzed the UK Biobank data and identified an excess of recent coalescence events in the immunoglobulin complex, which is attributable to positive selection during the past 20,000 years. While this signal is centered telomerically to IGHG3, the resolution of the method is of the order of megabases and may imply that the genomic region here considered is also involved.…”

An approximate full-likelihood method for inferring selection and allele frequency trajectories from DNA sequence data

“…A common 69 strategy for obtaining the full likelihood has been to find the distribution of the genealogy under 70 selection. For example, Krone and Neuhauser described the distribution of the coalescence tree of 71 a locus under weak selection and no recombination [27]. Alternatively, one can describe how the 72 genealogy depends on the trajectory of the selected allele (first described by [28]), and in turn how 73 2/46 the trajectory depends on selection.…”

“…It may also possible to make use of recent advances in inferring pairwise coalescence times 613 (e.g., [71]) to build an approximation to the full likelihood. Recently, Albers & McVean proposed 614 a composite likelihood method to estimate allele age by "sandwiching" the age using identity-by-615 descent tracts at the site of interest [72].…”