Mutation: lessons from RNA models
MetadataShow full item record
Mutation is a fundamental process in evolution because affects the amount of genetic variation in evolving populations. Molecular-structure models offer significant advantages over traditional population-genetics models for studying mutation, mainly because such models incorporate simple, tractable genotype-to-phenotype maps. Here, I use RNA secondary structure models to study four basic properties of mutation. The first section of this thesis studies the statistical properties of beneficial mutations. According to population genetics theory, the fitness effects of new beneficial mutations will be exponentially distributed. I show that in RNA there is sufficient correlation between a genotype and its point mutant neighbors to produce non-exponential distributions of fitness effects of beneficial mutations. These results suggest that more sophisticated statistical models may be necessary to adequately describe the distribution of fitness effects of new beneficial mutations. The second section of this thesis addresses the dynamics of deleterious mutations in evolving populations. There is a vast body of theoretical work addressing deleterious mutations that almost universally assumes that the fitness effects of deleterious mutations are static. I use an RNA simulation model to show that, at moderately high mutation rates, initially deleterious mutations may ultimately confer beneficial effects to the individuals harboring them. This result suggests that deleterious mutations may play a more important role in evolution than previously thought. The third section of this thesis studies the global patterns of mutations connecting phenotypes in fitness landscapes. I developed a network model to describe global characteristics of the relationship between sequence and structure in RNA fitness landscapes. I show that phenotype abundance varies in a predictable manner and critically influences evolutionary dynamics. A study of naturally occurring functional RNA molecules using a new structural statistic suggests that these molecules are biased towards abundant phenotypes. These results are consistent with an "ascent of the abundant" hypothesis, in which evolution yields abundant phenotypes even when they are not the most fit. The final section of this thesis addresses the evolution of mutation rates infinite asexual populations. I developed an RNA-based simulation model in which each individual's mutation rate is controlled by a neutral modifier locus. Using this model, I show that smaller populations maintain higher mutation rates than larger populations. I also show that genome length and shape of the fitness function do not significantly determine the evolved mutation rate. Lastly, I show that intermediate rates of environmental change favor evolution of the largest mutation rates.