Computational modeling of interactomes: Tools, techniques and application to genomes of apicomplexan parasites
Feb 13, Tue 2007
11:00am - 12:00pm, 2001 Mallott
Dr. Shailesh Date
Dept. of Genetics, University of Pennsylvania School of Medicine
Genome-scale reconstruction of protein-protein interaction networks proves an efficient strategy for gathering functional information about proteins which cannot be characterized using homology-based methods alone. By examining linkages between characterized and unknown (or hypothetical) entities, individual proteins or logical protein groups can be assigned putative function based on the ‘guilt-by-association’ principle, within the context of the network. Such large-scale interactome models also provide information about local and global relationships between gene products, and can be used to elucidate novel pathways, study cellular behavior, or predict outcomes based on possible propagation routes of effects exerted by small molecules.
We recently developed a schema that allows network reconstruction by integrating computational and experimental functional genomics data within a Bayesian framework. Application of this schema to the genome of the malarial parasite Plasmodiumfalciparum, the causal organism of the deadliest form of the disease in humans, generated a network that covered ~68% of the proteome, and provided functional information for more than 2000 uncharacterized proteins. We are now extending this work to investigate proteomes of two other apicomplexan parasites- Cryptosporidiumparvum and Toxoplasma gondii, as well as those of humans and mice. Information obtained from these analyses will be used to study common components and conserved interactions in apicomplexa, and will also be used to generate a framework for studying systems and pathways that play an important role in host-pathogen interactions.