Rapid identification and interpretation of gene–environment associations using the new R.SamBada landscape genomics pipeline

Susana Dunner,Estelle Rochat,Stéphane Joost,Elia Vajana,Michael W Bruford,Natalia Sevane,Oliver Selmoni,Sylvie Stucki,Pablo Orozco‐Terwengel,Kevin Leempoel,Solange Duruz

doi:10.1111/1755-0998.13044

Abstract

samβada is a genome–environment association software, designed to search for signatures of local adaptation. However, pre‐ and postprocessing of data can be labour‐intensive, preventing wider uptake of the method. We have now developed R.SamBada, an r‐package providing a pipeline for landscape genomic analysis based on samβada, spanning from the retrieval of environmental conditions at sampling locations to gene annotation using the Ensembl genome browser. As a result, R.SamBada standardizes the landscape genomics pipeline and eases the search for candidate genes of local adaptation, enhancing reproducibility of landscape genomic studies. The efficiency and power of the pipeline is illustrated using two examples: sheep populations from Morocco with no evident population structure and Lidia cattle from Spain displaying population substructuring. In both cases, R.SamBada enabled rapid identification and interpretation of candidate genes, which are further discussed in the light of local adaptation. The package is available in the r CRAN package repository and on GitHub (github.com/SolangeD/R.SamBada).

Highlights

Local adaptation implies the existence of advantageous alleles conferring a population living in its native habitat a higher fitness than any other allochthonous population living in the same habitat (Kawecki & Ebert, 2004)
Population structure can be accounted for by treating one or several population variables as environmental variables in multivariate analysis. samβada is written in C++ with a particular emphasis on high‐performance computing (HPC)
The first data set consists of 160 Moroccan sheep genotyped with whole genome sequencing (WGS) and characterized by no clear population structure, while the second one encompasses a Spanish Lidia Cattle population of 349 samples genotyped with 50 K single‐nucleotide polymorphisms (SNPs) chip, with one population variable

Summary

| INTRODUCTION

Local adaptation implies the existence of advantageous alleles conferring a population living in its native habitat a higher fitness than any other allochthonous population living in the same habitat (Kawecki & Ebert, 2004). The first data set consists of 160 Moroccan sheep genotyped with whole genome sequencing (WGS) and characterized by no clear population structure, while the second one encompasses a Spanish Lidia Cattle population of 349 samples genotyped with 50 K SNP chip, with one population variable. Both data sets are already published (see Data availability section) but have not yet been analysed with samβada

| MATERIALS AND METHODS

Findings

| DISCUSSION