Exome-wide DNA capture and next generation sequencing in domestic and wild species

Abstract

Background: Gene-targeted and genome-wide markers are crucial to advance evolutionary biology, agriculture, and biodiversity conservation by improving our understanding of genetic processes underlying adaptation and speciation. Unfortunately, for eukaryotic species with large genomes it remains costly to obtain genome sequences and to develop genome resources such as genome-wide SNPs. A method is needed to allow gene-targeted, next-generation sequencing that is flexible enough to include any gene or number of genes, unlike transcriptome sequencing. Such a method would allow sequencing of many individuals, avoiding ascertainment bias in subsequent population genetic analyses.We demonstrate the usefulness of a recent technology, exon capture, for genome-wide, gene-targeted marker discovery in species with no genome resources. We use coding gene sequences from the domestic cow genome sequence (Bos taurus) to capture (enrich for), and subsequently sequence, thousands of exons of B. taurus, B. indicus, and Bison bison (wild bison). Our capture array has probes for 16,131 exons in 2,570 genes, including 203 candidate genes with known function and of interest for their association with disease and other fitness traits.

Results: We successfully sequenced and mapped exon sequences from across the 29 autosomes and X chromosome in the B. taurus genome sequence. Exon capture and high-throughput sequencing identified thousands of putative SNPs spread evenly across all reference chromosomes, in all three individuals, including hundreds of SNPs in our targeted candidate genes.

Conclusions: This study shows exon capture can be customized for SNP discovery in many individuals and for non-model species without genomic resources. Our captured exome subset was small enough for affordable next-generation sequencing, and successfully captured exons from a divergent wild species using the domestic cow genome as reference.

Figure 1

Proportion of targeted exonic base pairs with a consensus genotype. All have a Phred-like quality score of at least 30. Total base pair counts, in millions, are plotted at selected minimum depths of coverage. Our estimates of exonic fixed differences and SNPs are based on consensus genotypes with coverage of at least 8 ×.

Figure 2

Chromosomal positions. a, all consensus base differences from the reference taurine genome, and b, heterozygous SNPs only. Both maps show consensus bases with at least 8 × coverage and Phred-like quality of at least 30. Numbers in the legends give totals for variants for each species.