Population Genomics and Quantitative Traits of Domestics and Wild Yeast

Recent resequencing efforts have advanced our understanding on genome evolution and population structure of S. cerevisiae and S. paradoxus. Although S. cerevisiae and S. paradoxus are closely related, they can be regarded as two distinct model organisms. S. cerevisiae has a strong association with human activity and the domestication of some strains has been proposed. However, S. paradoxus is thought to have had little contact with human activity. From the genome sequence analysis we find a clear population structure in S. paradoxus where sequence divergence strongly correlates with geographic origin. In contrast, S. cerevisiae strains are split into two groups: a well-defined geographic lineage of wild isolates, and a group of strains comprising baking, wine, and clinical isolates with mosaic genomes. Quantitative growth analysis of the sequenced strains exposed to over 200 environmental conditions revealed extreme phenotypic diversity and a clear separation among major lineages for specific traits. A current challenge is to dissect the genetic mechanisms underlying this natural phenotypic variation. Here, I will discuss how this population genomics resource can be used to pinpoint functional variants using linkage and association analysis in both natural and “artificial” Saccharomyces populations. A large fraction (31%) of the quantitative trait loci detected were mapped within subtelomeres, emphasising the importance of these regions in defining variation in natural populations and highlighting a current limit of genome projects where subtelomeric regions are underrepresented. Future studies will help elucidate the molecular mechanisms of causative segregating sites and how they vary within natural populations.