(513g) Dissecting the Genotype-to Phenotype Map in Eukaryotes: Molecular Determinants of Dominance, Heterosis, Pleiotropy, and Epistasis in Complex Traits

Breakthroughs in genome editing technology now allow the generation of designer organisms with user-specified nucleic acid content. But it is as yet unclear what genotype the user should encode. The intrinsic complexity of heritable traits has been known for over a century: Fisher assessed the statistical impact of polygenicity, dominance, heterosis, pleiotropy, and epistasis on heredity in 1918. Although the prevalence of these phenomena is clear in patterns of inheritance, the molecular variation that is responsible, such as changes in protein coding sequences or regulatory regions, remains obscure. To design organisms with optimal phenotypes, we must understand in great detail how genotypes map to expressed traits.

Using a panel of more than 18,000 fully genotyped F₆ diploid yeast derived from a cross between two wild Saccharomyces cerevisiae isolates, we identified thousands of genetic variants responsible for over a dozen complex traits. The unprecedented statistical power of the extra-large segregant panel allowed us to identify hundreds of genetic loci underlying each quantitative trait, frequently with single-nucleotide resolution. Moreover, we measured the precise molecular variation that underlies genetic nonlinearities and interactions. We developed a detailed picture of the diverse mechanistic contributors to complexity, pleiotropy, heterosis, dominance, and epistasis. Previously, these measurements have been made comprehensively only for the most radical type of genetic variant: the gene deletion. We found that, in fact, all types of natural molecular variation, including synonymous variants, make significant contributions to the genetic architecture of complex traits.

The genome design space, even for simple eukaryotes, is unimaginably vast: the genetic variation segregating in our panel alone allows for dramatically more possible genotypes than there are atoms in the universe. Understanding the architecture of complex traits at the level of single nucleotides will therefore be crucial for developing design rules for genome editing and engineering.