Genetic interactions underlie mechanisms of all heredity diseases. By revealing how one gene activity affects the function of another, genetic interaction studies complement other genomic approaches such as protein-protein interaction studies by providing direct evidence of functional relations among genes, offering answers to questions such as functional redundancy and genetic epistasis that other approaches can not resolve. Genetic interaction research becomes especially important when studying disease genes, because human genetic diseases often result from a combined action of multiple genes. My long-term research goal is to reach a system level understanding of the genetic interaction network. I am interested in combining my expertise in experimental biology and computer science to design high-throughput tools and apply these tools to study genetic interactions at genome scale. The objective of this application is to generate a large-scale, quantitative genetic interaction map of disease genes in the metazoan model C. elegans. To accomplish this, I have developed a high-throughput pipeline for predicting and testing of genetic interactions in C. elegans. The first component of the pipeline is a computational system that generates genome-wide probabilistic predictions of genetic interactions by integrating expression, phenotype, interaction, and function data from multiple species. The second component of the pipeline is a high-speed automatic phenotyping system that uses automatic microscopy and image processing to provide quantitative measurements of nematode phenotypes. I plan to improve this research pipeline by improving the quality of genetic interaction predictions, extending the predictions to other metazoans including human, and increasing the number of phenotypes that can be scored by the automatic phenotyping system. I will then apply the established tools to systematically investigate the interactions of human disease gene homologs in C. elegans. A multi-phenotypic, quantitative model of the disease gene interaction network will be generated. For each disease gene, I will identify its interacting partners, the type of interaction between each gene pair, and the biological process for each interaction. This study is relevant to public health because it brings insights to disease mechanisms and it provides an efficient system to identify new targets for preventive and therapeutic intervention of human heredity diseases. ? ? ?
