Understanding the drivers of speciation is critical to interpreting patterns of biodiversity. The identification of the genetic changes underlying adaptation and reproductive isolation is necessary to link barriers to gene flow to the causal origins of divergence.Here, we present a novel approach to the genetics of speciation, which should complement the commonly used approaches of quantitative trait locus mapping and genome-wide scans for selection. We present a large-scale candidate gene approach by means of sequence capture, applied to identifying the genetic changes underlying reproductive isolation in the pea aphid, a model system for the study of ecological speciation. Targeted resequencing enabled us to scale up the candidate gene approach, specifically testing for the role of chemosensory gene families in host plant specialization. Screening for the signature of divergence under selection at 172 candidate and noncandidate loci, we revealed a handful of loci that show high levels of differentiation among host races, which almost all correspond to odorant and gustatory receptor genes. This study offers the first indication that some chemoreceptor genes, often tightly linked together in the genome, could play a key role in local adaptation and reproductive isolation in the pea aphid and potentially other phytophagous insects. Our approach opens a new route toward the functional genomics of ecological speciation.

BackgroundThe advent and democratization of next generation sequencing and genotyping technologies lead to a huge amount of data for the characterization of population genetic diversity in model and non model-species. However, efficient storage, management, cross-analyzing and exploration of such dense genotyping datasets remain challenging. This is particularly true for the bovine species where many SNP datasets have been generated in various cattle populations with different genotyping tools.DescriptionWe developed WIDDE, a Web-Interfaced Next Generation Database that stands as a generic tool applicable to a wide range of species and marker types (http://widde.toulouse.inra.fr). As a first illustration, we hereby describe its first version dedicated to cattle biodiversity, which includes a large and evolving cattle genotyping dataset for over 750,000 SNPs available on 129 (89 public) different cattle populations representative of the world-wide bovine genetic diversity and on 7 outgroup bovid species. This version proposes an optional marker and individual filtering step, an export of genotyping data in different popular formats, and an exploration of genetic diversity through a principal component analysis. Users can also explore their own genotyping data together with data from WIDDE, assign their samples to WIDDE populations based on distance assignment method and supervised clustering, and estimate their ancestry composition relative to the populations represented in the database.ConclusionThe cattle version of WIDDE represents to our knowledge the first database dedicated to cattle biodiversity and SNP genotyping data that will be very useful for researchers interested in this field. As a generic tool applicable to a wide range of marker types, WIDDE is overall intended to the genetic diversity exploration of any species and will be extended to other species shortly. The structure makes it easy to include additional output formats and new tools dedicated to genetic diversity exploration.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-015-2181-1) contains supplementary material, which is available to authorized users.