Detection of Shared Balancing Selection in the Absence of Trans-Species Polymorphism

Abstract

Trans-species polymorphism has been widely used as a key sign of long-term balancing selection across multiple species. However, such sites are often rare in the genome and could result from mutational processes or technical artifacts. Few methods are yet available to specifically detect footprints of trans-species balancing selection without using trans-species polymorphic sites. In this study, we develop summary- and model-based approaches that are each specifically tailored to uncover regions of long-term balancing selection shared by a set of species by using genomic patterns of intraspecific polymorphism and interspecific fixed differences. We demonstrate that our trans-species statistics have substantially higher power than single-species approaches to detect footprints of trans-species balancing selection, and are robust to those that do not affect all tested species. We further apply our model-based methods to human and chimpanzee whole-genome sequencing data. In addition to the previously established major histocompatibility complex and malaria resistance-associated FREM3/GYPE regions, we also find outstanding genomic regions involved in barrier integrity and innate immunity, such as the GRIK1/CLDN17 intergenic region, and the SLC35F1 and ABCA13 genes. Our findings not only echo the significance of pathogen defense but also reveal novel candidates in maintaining balanced polymorphisms across human and chimpanzee lineages. Finally, we show that these trans-species statistics can be applied to and work well for an arbitrary number of species, and integrate them into open-source software packages for ease of use by the scientific community.

Learning the properties of adaptive regions with functional data analysis

“…Moreover, our SURFDAWave approach is not restricted to application on adaptive introgression and selective sweep scenarios, and can be implemented for probing genomic variation of other evolutionary processes that leave a spatial or temporal signature in genomic data. Such examples include the identification of genomic targets of balancing selection [ e.g ., 54 , 55 , 56 , 57 , 58 , 59 ], complex forms of adaptation such as staggered selective sweeps [ 60 ] that have yet to be interrogated for in genomic data, and non-adaptive processes such as recombination rate estimation [ e.g ., 8 , 7 , 61 ].…”

Flexible Mixture Model Approaches That Accommodate Footprint Size Variability for Robust Detection of Balancing Selection

“…2018 ; Ye et al. 2018 ; Cheng and DeGiorgio 2019 ). However, despite multiple efforts to design statistics for identifying balanced loci ( DeGiorgio et al.…”

Three hallmarks of malaria-induced selection in human genomes

“…As a result, scans for balancing selection fail to identify either of these genes as outliers (e.g. Andrés et al 2009; DeGiorgio et al 2014; Bitarello et al 2018), unless they incorporate non-human polymorphism (Leffler et al 2013; Chang and DeGiorgio 2019). However, intraspecies haplotype structure alone may convey a signal of selection if tested for directly (Figure 1A).…”

“…Thus, despite being canonical examples of adaptive polymorphism, these genes are almost never detected in genome-wide scans for partial sweeps or balancing selection (e.g. Voight et al 2006; Akey 2009; Andrés et al 2009; Leffler et al 2013; DeGiorgio et al 2014; Siewert and Voight 2017; Bitarello et al 2018; Cheng and DeGiorgio 2019).…”

“…Genome-wide scans for selection in humans are now routine (Fan et al 2016), but they have poor replicability such that different methods produce very different lists of candidate genes. This is true for positive selection (Akey 2009) and may be worse for balancing selection: among six recent studies that scan the genome for balancing selection in Africans or African Americans, the proportion of identified candidate selection targets that are shared between any two scans ranges from 0 to 9% (Andrés et al 2009; Leffler et al 2013; DeGiorgio et al 2014; Siewert and Voight 2017; Bitarello et al 2018; Cheng and DeGiorgio 2019). For some genes, selection has been effectively validated phenotypically because allelic effects on infection or fitness have been demonstrated.…”