DMS/NIGMS 2: Building and Testing a Graph Theory Toolbox for Uncovering Redundant Pathways in Genetic Interaction Networks - Biological networks are powerful resources for the discovery of genes and genetic modules. For the classical physical protein-protein interaction network, as well as for other types of pairwise association data between genes or proteins that is organized according to homophilic principles, diffusion-based low dimensional network embedding methods have proved quite powerful. While at a coarse scale, genetic interaction networks, built from high-throughput epistasis experiments, also display some homophily in their organization, at a fine scale, they display very different graph-theoretic structure that can be leveraged to find mechanisms of redundancy and fault tolerance among biological pathways active in the cell. However, the graph theoretic toolbox to analyze this fine-grained structure in genetic interaction networks is still underdeveloped, compared with algorithms developed for analysis of the purely homophilic biological networks. Inspired by some of the tools and techniques used in graph theory to study vertex cuts in networks, we propose to develop a more formal framework for categorizing the patterns of resilience and redundant pathway mechanisms in genetic interaction networks, and new algorithms for discovering genes involved in compensatory pathways. We will use our mathematical framework to computationally make phenotype predictions for the results of gene knockout experiments with either unseen gene combinations or in unseen environmental conditions. We will validate some of our predictions focusing on alternative pathways for DNA replication and repair in Saccharomyces cerevisiae (baker's yeast). Since DNA replication and repair are crucial pathways for cancer cell survival, identification of genes in those subnetworks could reveal new cancer therapy targets. Discovery of new genes is important for a basic understanding of cell functioning and to elucidate the cause of inherited genetic disease. RELEVANCE (See instructions): Saccharomyces cerevisiae is a foundational model system for studying mechanisms of resilience and fault tolerance in the Eukaryotic cell, and the pathways for DNA replication and repair are sufficiently evolutionarily conserved to make insights in these model systems relevant for understanding human diseases that exhibit genome instability. Methods to target compensatory pathways are also an active area of drug design, including in the search for personalized therapies customized by tumor in cancer.
