Population Genomics of Saccharomyces Yeasts

Abstract

This thesis examines genome-wide polymorphisms amongst 20 strains of Saccharomyces paradoxus, a yeast strain which has recently emerged as a model organism for population genetic studies. Three major studies are included in this thesis. The first study attempts to quantify the life cycle of yeast undergoing different modes of reproduction in nature. Measures of mutational and recombinational diversity are used to make two independent estimates of the population size. In an obligatory sexual population these estimates should be approximately equal. Instead, there is a discrepancy of about three orders of magnitude, indicating that S. paradoxus goes through one sexual cycle once in every ~1,000 asexual generations. This study illustrates the utility of population genomic data in quantifying the life cycles of organisms undergoing different modes of reproduction. Second, a map showing the distribution of rates of population recombination along chromosome III of S. paradoxus is presented. Several regions of very high recombination (hotspots) are identified in chromosome III. Comparison of hotspot regions between S. paradoxus and S. cerevisiae shows evidence of conservation of recombination hotspot regions. I argue that these observations reflect the weak impact of recombination due to the reduced frequency of sex of yeasts in nature. Recombination rates correlate with GC content, consistent with various studies in yeasts and humans, but there is no correlation with diversity or divergence. In addition, regions of extremely high recombination (hotspots) show an increased rate of GC→AT than rest of the chromosome. The reason for this is not clear at present ii Finally, a catalogue of polymorphisms within each population, and divergences between the two populations of S. paradoxus is presented. Tests of selection on the chromosomal sequence of S. paradoxus suggest a predominant mode of purifying selection. At least a third of mutations in synonymous sites and ~90% of mutations in replacement sites are removed by purifying selection. We estimate that around 12-31% of replacement mutations are deleterious in S. paradoxus, and that the average selection strength acting on these mutations is 1%. I also present a summary of data and subsequent analyses from the Saccharomyces Genome Resequencing Project (SGRP). Population genetic measures are applied to data using different basecalling quality cutoffs. From the results I recommend that at least a quality score of 40 is necessary to achieve the confidence required in data to be used in population genomic analyses.