Finding the genes underlying complex traits is difficult. We show that new sequencing technology combined with traditional genetic techniques can efficiently identify genetic regions underlying a complex and quantitative behavioral trait. As a proof of concept we used phenotype-based introgression to backcross loci that control innate food preference in Drosophila simulans into the genomic background of D. sechellia, which expresses the opposite preference. We successfully mapped D. simulans introgression regions in a small mapping population (30 flies) with whole-genome resequencing using light coverage (1·). We found six loci contributing to D. simulans food preference, one of which overlaps a previously discovered allele. This approach is applicable to many systems, does not rely on laborious marker development or genotyping, does not require existing high quality reference genomes, and needs only small mapping populations. Because introgression is used, researchers can scale mapping population size, replication, and number of backcross generations to their needs. Finally, in contrast to more widely used mapping techniques like F 2 bulk-segregant analysis, our method produces near-isogenic lines that can be kept and reused indefinitely.
COMPLEX traits are inherently difficult to dissect genetically. Quantitative trait loci (QTL) studies, genome-wide association studies (GWAS), and forward and reverse genetics are all powerful tools; however, each technique has inherent weaknesses that limit ability to find causal loci (Mackay et al. 2009;Manolio et al. 2009). New methods using nextgeneration sequencing (NGS) technology have successfully captured single loci underlying Mendelian traits generated from mutagenesis screens (Sarin et al. 2008;Blumenstiel et al. 2009; Flibotte et al. 2010;Laitinen et al. 2010;Sarin et al. 2010;Xia et al. 2010;Zuryn et al. 2010); recombinant inbred lines (RILs) Schneeberger et al. 2009); and backcross populations using dominant markers (Andolfatto et al. 2011). However, we lack a time-and costeffective method that maps multiple loci simultaneously without a priori knowledge of their location, number, or effect size.Mapping complex traits is more challenging than mapping Mendelian traits. QTL studies in the past decades have uncovered a plethora of loci underlying complex traits, but QTL methods lack the power to resolve candidate regions to individual genes (Mackay et al. 2009). New NGS approaches have the potential to capture multiple causative loci; however, these methods may also lack sufficient power. proposed "whole-genome resequencing" (WGR), and Baird et al. (2008) proposed "restriction site associated DNA" (RAD) genotyping, which both use NGS-based mapping on bulk segregant populations (Michelmore et al. 1991). These methods require large mapping populations to detect multiple loci of weak effect (e.g., Ehrenreich et al. 2010). Andolfatto et al. (2011) developed a "multiplex shotgun genotyping" (MSG) method, yet MSG also relies on large backcross populations, and i...