Patterns of crossing over and gene conversion in meiotic recombination

Abstract

Recombination is an essential component of meiosis, which acts to shuffle genetic variation within the genome, and is required for proper chromosomal disjunction. Recombination has two outcomes: crossover and gene conversion. Crossover consists of an equal exchange of genetic material between chromosomes, while gene conversion is the non-reciprocal transfer of genetic information between homologues that occurs over smaller intervals. Most research has focused on crossover, which varies widely in frequency and placement within the genome, and between individuals, sexes, populations, and species.;In this thesis, I have employed statistical methods to identify sex specific differences in recombination in detail. In humans, I performed a pedigree analysis using data derived from over 18,000 meioses. I found that males have a higher proportion of crossovers overlapping hotspots than females, a 4.6% increase. In addition, I measured crossover interference, which affects the spatial positioning of crossovers, finding that older mothers had a steep increase in crossovers that escape regulation by interference. These crossovers appear closely spaced, pointing to a possible deregulation of recombination that increases with maternal age.;I also examined sex differences in recombination using a complex pedigree of inbred dogs, consisting of 408 meioses. Dogs are unique in that their PRDM9 ortholog has accumulated mutations, rendering it inactive, and raising questions as to how crossovers are placed without this protein, which is known to specify the locations of hotspots in the genome. I found that dog recombination is broadly similar to that of humans. Dog recombination appears to be more concentrated to a smaller proportion of sequence when compared to humans, suggesting the existence of hotspots. I found evidence for positive crossover interference acting in the dog genome with a similar mechanism to that observed in humans.;In addition, I developed a novel computational approach using a hidden Markov model to detect gene conversion in population genetic data. The model uses two divergent ancestral reference populations in order to predict the location of gene conversion events on an admixed haplotype. Simulation results show that the model can plausibly be used to detect gene conversion.;Overall, I have performed three separate but interrelated studies of recombination that expand on existing work. These studies come together to enable further understanding of the properties of genetic recombination as a whole.