Evolution and natural selection
JV Chamary
Under construction. I study the evolution of mammalian genomes, with a particular interest in the role that selection has played in shaping DNA architechture and composition.
Copyright 01.02.07 JVC
Patterns and processes of evolution at silent sites in mammalian genes
Within genes, mutations that do not change proteins are usually considered 'silent' because it is assumed that they have no affect on an organism's phenotype. Evolution at silent sites is expected to be selectively neutral, so that the probability that a new mutation will become fixed in a population is dictated by chance. This is especially important in mammals, where small effective population sizes should reduce the ability of natural selection to influence the fate of mutations with presumed negligible impacts on fitness, such as silent nucleotide substitutions. If true, analysis of silent sites would provide a window into the process of mutation. Despite claims to the contrary, I show that neighbouring genes have similar rates of silent evolution. If silent sites evolve neutrally, this pattern reflects regional variation in the mutation rate across the genome. But are silent sites neutrally evolving? I provide multiple lines of evidence that suggest not. In rodents, although two classes of silent sites, introns and synonymous sites, evolve at similar rates, the process of evolution is different between the two. After eliminating conserved intronic sequence, I observe a preference for cytosine at synonymous third sites of codons. I then investigate two models by which synonymous sites might be functional. First, in the mouse lineage, the cytosine preference at third sites and the avoidance of mutations at some synonymous sites can potentially be explained by selection on mRNA secondary structure to promote optimal stability. Second, in humans, biases for particular codons over their synonyms increases near intron-exon junctions, which largely reflects the presence of exonic splicing enhancers. I show that the effect of purifying selection on putative enhancers in mammals leads to a reduction in estimates of the mutation rate.