How might the process of meiotic crossing-over affect the evolution of sequences and genome structure? Much attention has been focussed on the notion that crossing-over modulates the efficacy of selection. Here, we consider how good the evidence for this is. Correlations between recombination and protein rates of evolution, commonly interpreted in the above framework, might be misleading for failing to remove the effects of covariates or misinterpreted by disregarding direct effects of recombination, such as biased gene conversion. Similarly, higher diversity commonly seen in highly recombining domains may not necessarily imply a connection between recombination and diversity. It could, for instance, reflect a covariance with mutation rate variability owing to replication timing effects. This thesis examines not only these links betweenrecombination and gene evolution but in addition asks other recombination-centred questions. Is gene order evolution in part driven by recombination predisposing certain sites to rearrangement? How can we account for the genomic location of recombination? Does, for example, germline expression, recently suggested to predict low recombination rates in mammals, predict recombination rates in flies? With the exception of the second chapter where we investigate the relationship between double strand break formation, recombination and sequence divergence in Saccharomyces cerevisiae, my work considers Drosophila, making use of the 12 genomes resource.While an effect of crossover on divergence owing to more efficient selection cannot be ruled out, we demonstrate that premeiotic double strand breaks also predict slow evolution. Late replication, we show, is associated with increases in both divergence and variation but this does not undermine the recombination-diversity correlation. While recombination is associated with increased rearrangement rates, we find no evidence that germline expressed genes avoid recombination.
|Date of Award||1 Nov 2011|
|Supervisor||Laurence Hurst (Supervisor)|