UCLA

Most, if not all, institutions benefit from diverse populations. The benefits of diversity are also seen in non-human populations, including on the levels of genomes. My three dissertation projects all touch on processes that can influence diversity in some capacity. My first chapter describes the effectiveness of the University of California, Los Angeles (UCLA) Competitive Edge (CE) bridge program in supporting PhD students from historically excluded and underrepresented groups (URG). Through surveys of 55+ first-year students, my study reveals CE's success in enhancing key aspects such as mentoring relationships, socialization, and overall preparedness. These are all crucial factors influencing student retention. Moving into the genetic sphere, my second chapter examines meiotic recombination, a process generating genetic diversity in sexually reproducing species through the decoupling of alleles. Focusing on the impact of domestication on recombination rates, my study uses wolves and breed dogs as a model. I tested and rejected the hypothesis that domestication leads to increased recombination rates. My work shows intriguing patterns emerge. For example, border collies exhibit higher inferred recombination rates, while pugs have lower rates compared to wolves. Despite these differences, I estimated a stable recombination landscape in dogs and wolves. Additionally, my work created a genetic map for further exploration of the canid genome. My final chapter employs this genetic map to investigate the role of recombination in inferring the distribution of fitness effects (DFE) of new mutations. The DFE is a crucial aspect in population genetics because it reveals how selection impacts genetic diversity. However, Poisson Random Field (PRF) methods of DFE inferences assume sites are unliked. My work investigates how estimates of the DFE vary with linkage and recombination rates in wolves. I find that estimating the DFE in low recombination regions is similar to high recombination regions, despite differences in patterns of linked selection. Thus, my results suggest that DFE inference using PRF methods is not significantly biased by linked selection. Altogether, these results have implications for mechanisms that influence diversity in their relevant contexts.