Exploiting natural selection to study adaptive behavior

The research presented in this dissertation explores different computational and modeling techniques that combined with predictions from evolution by natural selection leads to the analysis of the adaptive behavior of populations under selective pressure. For this thesis three computational methods were developed: EXPLoRA, EVORhA and SSA-ME. EXPLoRA finds genomic regions associated with a trait of interests (QTL) by explicitly modeling the expected linkage disequilibrium of a population of sergeants under selection. Data from BSA experiments was analyzed to find genomic loci associated with ethanol tolerance. EVORhA explores the interplay between driving and hitchhiking mutations during evolution to reconstruct the subpopulation structure of clonal bacterial populations based on deep sequencing data. Data from mixed infections and evolution experiments of E. Coli was used and their population structure reconstructed. SSA-ME uses mutual exclusivity in cancer to prioritize cancer driver genes. TCGA data of breast cancer tumor samples were analyzed.status: publishe