A quantitative examination of cellular mechanisms that modulate the impacts of genetic variation - Renewal - 1

PROJECT SUMMARY Predicting the phenotypic impacts of a mutation is a major goal in biology and medicine. But the paths linking genotype to phenotype are difficult to navigate. For one, some phenotypes impact others, so the impacts of mutation can stretch out across networks of related traits. One of my lab’s goals is to investigate the relationships between basic features of cells (e.g., misfolded protein abundance, levels of protein-folding chaperones, and cell growth rate) so we can predict some phenotypes from others. But doing so is not enough. Predicting phenotype is more challenging than this because the impacts of mutation, and the networks of related traits through which they spread, can change across contexts. By re-measuring the relationships between traits in many different genetic backgrounds and environments, my lab endeavors to make headway on one of the major goals of modern biology: predicting the phenotypic impacts of mutation in diverse contexts. To achieve this goal, my lab conducts high-throughput experiments in the model eukaryote, budding yeast, that simultaneously quantify the phenotypic impacts of many mutations across many environments. We interpret these big datasets using diverse mathematical models and machine learning approaches. In some projects, we quantify the correlations between phenotypes to infer the network through which a mutation’s influence travels and how that network changes across contexts. For example, I recently measured the correlations between yeast single-cell morphological traits, how they change across the cell cycle, and how this predicts which traits are jointly influenced by mutation. In the next five years, my lab plans to apply a similar strategy to study how the impacts of mutation travel through a regulatory network. In other projects, we deconvolute high- dimensional data into an abstract genotype-phenotype map that uses shared mutant behavior across contexts to improve fitness predictions. In the next five years, we plan to apply this approach to make better predictions about how the fitness of drug-resistant yeast mutants changes across subtly different concentrations and combinations of drugs. Some of our work involves engineering libraries of mutant yeast strains to investigate scaling relationships describing how molecular-level phenotypes (e.g., the levels of misfolded proteins or ribosomes) impact higher- level properties (e.g., cell growth and fitness). Other work focuses on large collections of adaptive mutations generated by laboratory evolution experiments. Our overall goal is to build predictive maps from genotype to phenotype and, in so doing, to learn about molecular biology, and also generate tools to measure and predict some phenotypes from others that will be broadly useful to the community.