Individuals of the same species exhibit extensive phenotypic variation. Although there is a large genetic component to this phenotypic diversity, the molecular basis of natural phenotypic variation is not well understood. For example, is this diversity due to many or a few DNA sequence polymorphisms? Do these variants tend to lie in protein coding or gene regulatory sequences? What are the effects of these polymorphisms in different genetic backgrounds? And what are the molecular interactions that underlie the complex genetic interactions between variants that control natural phenotypic variation. The goals of this project are to approach these questions by elucidating the molecular architecture of a complex, polygenic trait in the yeast Saccharomyces cerevisiae. Wild isolates of yeast show large differences in the efficiencies at which they sporulate (undergo meiosis) in response to carbon and nitrogen starvation. Using Quantitative Trait Loci (QTL) analysis, polymorphisms that contribute to this natural variation will be identified. An array of genetic, genomic, and computational tools will be deployed including microarray-based genome-wide SNP typing, genome-wide expression profiling, high-throughput flow cytometry, and a suite of computational algorithms designed to locate functional sequences through comparative sequence analysis. The range of tools available for this study should facilitate the identification of sequence polymorphisms that contribute to variation in sporulation efficiency. Successful completion of the aims of this project will provide a better understanding of the molecular architecture of natural variation and the role that changes in gene expression play in producing new phenotypes. Such an understanding is key to understanding abnormal variation, because in many (most?) cases it will be due to "bad" combinations of common alleles. These studies promise to inform our understanding of the molecular bases of phenotypic variation among individuals of a species, and will shed light on the mechanisms underlying the evolution of new phenotypes.
A large amount of the observable differences between individuals is the result of differences in their genes. In theory it is therefore possible to predict certain characteristics of an individual by reading the DNA sequence of the appropriate genes from that individual. In practice this is impossible for the vast majority of traits we care about. This is because in nature observable variation has a very complex genetic basis. It results from variations in multiple genes and those variations interacting with each other to produce a huge number of possible outcomes. Modern technological breakthroughs, especially in the genetically tractable bakers' yeast, promise to speed the identification of gene variants that contribute to observable differences between individuals in a population. The goal of this project is to characterize the genetic variations that cause differences between isolates of bakers' yeast to an extent such that the characteristics of strains can be accurately predicted from their DNA sequence alone.
The PI is active in promoting and improving the interdisciplinary training of graduate students at Washington University. He designed and is currently teaching a graduate level class called Introduction to Genomics, in which students are taught the principles and issues of a wide variety of approaches and techniques used in genomics, including genome mapping and sequencing, sequence similarity searches, computational gene finding, microarray studies, mass spectrometry, and the basic statistical and probabilistic principles necessary to understand functional genomics. All students also learn the Perl programming language and write their own software for parsing Blast reports, quantitating microarray data, building phylogenetic trees, clustering expression data, calculating linkage disequilibrium between SNP markers, and determining haplotype block structures from high-throughput genotyping data. The course has become part of the core curriculum for the Molecular Genetics program at Washington University. Many past students with no previous formal training in computational analyses have gone on to write their own scripts for their thesis projects.
