Advisor:
Dr. Patrick McGrath, School of Biological Sciences, Georgia Institute of Technology
Committee Members:
Dr. Greg Gibson, School of Biological Sciences; Georgia Institute of Technology
Dr. Todd Streelman, School of Biological Sciences; Georgia Institute of Technology
Dr. Annalise Paaby, School of Biological Sciences; Georgia Institute of Technology
Dr. Peng Qiu, School of Biomedical Engineering; Georgia Institute of Technology
Abstract:
Different species occupy different ecological niches, relying on a number of biological differences to maximize their chance for survival. These differences are mostly genetic, as evolution has fixed a large number of causal genetic variants through positive selection and genetic drift. Unfortunately, reproductive barriers prevent us from mapping the causal variants responsible for macroscopic evolution. In my thesis, I will use two systems, the C. elegans nematode and the Lake Malawi cichlids, that allow us to study evolution over long time scales. Previous studies of the McGrath lab have identified hundreds of regions of the C. elegans genome under long-term balancing selection. Mechanistic studies of one of these loci indicated that cis-regulatory differences between the two haplotypes were responsible for different foraging strategies that are optimized for different food conditions. To systematically study cis-regulatory evolution for the entire genome, I will first use an F1 hybrid RNA sequencing approach to identify allele specific expression (ASE) differences between two strains. I will also use cell specific sequencing to map these ASE back to individual neuron classes, focusing on sensory neurons. Secondly, I will study a locus under balancing selection that contains a selfish genetic incompatible set peel-1/zeel-1. I will use simulation and experimental assays to determine how this locus could be under balancing selection. Unexpectedly, my initials results indicate the selfish gene peel-1 is advantageous, suggesting that it also acts unselfishly. Finally, Cichlid species are unique in that they can create fertile cross-progeny, allowing for genetic mapping. To facilitate functional variant discovery between different species. We have developed a pipeline to automatically annotate cichlids behavior from videos. This pipeline uses a 3D convolutional neural network with HMM based focusing. Taken together, our novel approaches provide another entry point for functional variant studies in evolution.
