Automated prediction of protein complex substructure from genomics and proteomics data

点击翻译按钮获取中文摘要

Experimental determination of interactions among cellular components, including biochemical reactions, protein-protein and genetic interactions, charts a map of the cell. This map defines gene function and is the basis of our models of biological processes. This knowledge will eventually enable us to predict phenotype from genotype, understand the genetic basis of disease and engineer new biological systems for industrial applications. Recent biotechnology advances and the availability of completely sequenced genomes for hundreds of organisms have led to genome-scale mapping of networks of interacting genes, their RNA and protein products and small molecules. For instance, over 145,000 protein-protein, protein-DNA and genetic interactions for the yeast Saccharomyces cerevisiae have been mapped using techniques such as yeast two-hybrid, chromatin immunoprecipitation on microarrays and immunoprecipitation coupled to mass spectrometry, with significant contributions by Canadian research groups. Further, molecular profiling methods, such as RNA level measurement by microarray or protein quantitation by mass spectrometry, provide insight into the dynamic levels of cellular components. This extensive cellular network information demands the development of powerful computational tools to facilitate data integration and interpretation outside the direct capacity of the human brain. I aim to address this challenge by developing novel biological network visualization and analysis methods to discover new information about protein complexes. These methods will help us, as biologists, effectively navigate the cell map at multiple levels of detail and use it to make new discoveries.