Genomic characterization of simple and complex traits in non-model organisms

Abstract

The over-arching goal of my dissertation is to understand the relationship between genotypic and phenotypic variation in non-model organisms. I am interested in how selection affects gene networks leading to phenotypic variation for simple and complex traits and how past demographic events obscure genomic signatures of selection. My work mainly focuses on the Drosophila yakuba clade that encompasses D. yakuba, D. santomea, and D. teissieri.

I first propose a streamlined variant-calling pipeline for genotyping using high-coverage resequencing data and a simulation pipeline for identifying appropriate filters for any model or non-model organism. Using high quality variant calls, I found that of the D. santomea abdominal pigmentation genes: Abd-B, ebony, Pdm3, tan and yellow, only ebony has fixed differences in the amino acid sequence. I further confirmed the presence of a fixed transposable element (TE) insertion in the 5’ regulatory region of ebony in D. santomea but found no signatures of selection surrounding this TE. Next, I show that the unexpected discovery of long runs of homozygosity (ROH) tracts throughout the genomes of outbred D. santomea is most consistent with a demographic model with many small subpopulations. Lastly, I mapped causal SNPs for a trait with a binary phenotype: red – green gall color variation in aphids. I resolved the causal region in Hormaphis cornu to a 501bp region but homologous genes in the closely related species H. hamadelidis was not associated to gall coloration.