Abstract
As DNA sequencing throughput increases, novel strategies for discovering genes that affect traits of interest become available. One strategy starts with a population of animals and selects individuals over multiple generations for a particular trait. Subsequent whole genome sequencing should identify loci affecting this trait. We apply this strategy by sorting flies for wing length over 18 generations, obtaining two populations that differ in wing length by 20%. Flies with longer wings had increased overall body sizes and elevated TOR activity, suggesting that genetic variation targets TOR signaling to influence body size. High-throughput sequencing of big and small flies identified thousands of single nucleotide polymorphisms that differed between the two populations, leading us to identify five novel regulators of TOR signaling. Surprisingly, stochastic simulations of the process show that large fractions of the genetic differences between the big and small flies are probably biological false positives, selected by chance by random drift. We employ these computer simulations to identify experimental setup parameters to improve the signal-to-noise ratio for successfully running sort-and-sequence experiments – a resource which will hopefully be useful for the community.