Detecting Genomic Signatures of Natural Selection with Principal Component Analysis: Application to the 1000 Genomes Data

Abstract

To characterize natural selection, various analytical methods for detecting candidate genomic regions have been developed. We propose to perform genome-wide scans of natural selection using principal component analysis (PCA). We show that the common FST index of genetic differentiation between populations can be viewed as the proportion of variance explained by the principal components. Considering the correlations between genetic variants and each principal component provides a conceptual framework to detect genetic variants involved in local adaptation without any prior definition of populations. To validate the PCA-based approach, we consider the 1000 Genomes data (phase 1) considering 850 individuals coming from Africa, Asia, and Europe. The number of genetic variants is of the order of 36 millions obtained with a low-coverage sequencing depth (3×). The correlations between genetic variation and each principal component provide well-known targets for positive selection (EDAR, SLC24A5, SLC45A2, DARC), and also new candidate genes (APPBPP2, TP1A1, RTTN, KCNMA, MYO5C) and noncoding RNAs. In addition to identifying genes involved in biological adaptation, we identify two biological pathways involved in polygenic adaptation that are related to the innate immune system (beta defensins) and to lipid metabolism (fatty acid omega oxidation). An additional analysis of European data shows that a genome scan based on PCA retrieves classical examples of local adaptation even when there are no well-defined populations. PCA-based statistics, implemented in the PCAdapt R package and the PCAdapt fast open-source software, retrieve well-known signals of human adaptation, which is encouraging for future whole-genome sequencing project, especially when defining populations is difficult.

Genomic response to natural selection within alpine cattle breeds

“…The detection of genomic regions affected by natural selection was carried out based on the approach adopted in R package PCAdapt (Duforet-Frebourg et al 2016) according to Luu et al (2016). To identify the signals of selection, the Mahalanobis distance test statistic as a multivariate method measuring the distance of the point from the mean, was used.…”

“…In order to provide a list of variants that are potentially involved in the natural or artificial selection, genome scans measure the genetic differentiation between populations considering that extreme values correspond to candidate regions. Although high levels of differentiation could have various causes, the adaptation of individuals to their local environment is a prominent explanation to such patterns of differentiation for adaptive loci exceeding neutral expectations (Duforet-Frebourg et al 2016). …”

“…The genome scan based on F ST can produce many false positives of selection signature signals due to the various biological and statistical reasons. There are important caveats with approaches related to F ST because they require grouping individuals into populations, sometimes subjectively, resulting in the loss of important selection signals (Duforet-Frebourg et al 2016). Consequently, Duforet-Frebourg et al (2014) proposed an alternative approach to perform genome-wide scans of natural selection using principal component analysis (PCA) that uses multivariate evaluation for the population structure identification.…”

“…The PCA based statistic implemented in PCAdapt R package provides three main advantages compared to the classic F ST approach. It works on individual basis, the computation time is reduced in comparison to methods using the MCMC algorithms and candidate loci can be related to the different evolutionary events, which correspond to different principal components (PCs) (Duforet-Frebourg et al 2016;Luu et al 2016).…”

Molecular signatures of host specificity linked to habitat specialization in Exaiptasia sea anemones

“…Under a model of divergence among three populations, in which two populations share more recent evolutionary history compared to a third , outlier loci strongly correlated with principal components that separate these populations may be promising targets of differential selection pressure (Duforet‐Frebourg, Luu, Laval, Bazin, & Blum, 2016). To identify loci in E. pallida most strongly correlated with divergence between the Bocas‐specific and global populations, we performed outlier detection using principal component analysis (Luu et al., 2017).…”