Tools for Analyzing, Modeling, and Comparing Protein-Protein Interaction Networks. Start date: Jan. 15, 2007. Expires: Dec. 31, 2007. Awarded amount to date: $141,643. Principal investigator: Natasa Przulj. Sponsor: University of California-Irvine.
Supports a project to improve tools for analyzing and modeling protein-protein interaction networks. According to the grantees, currently available PPI networks “are partial and mostly resulting from various high-throughput proteomics experiments that attempt to discover as many PPIs as possible in a number of organisms.” The funded project introduces new, more sensitive measures of network local structure that are based on enumeration of “graphlets” — small induced subgraphs of large graphs. Preliminary results indicate that geometric random graphs provide a “superior fit” to the currently available PPI networks than do other network models, including scale-free networks.
Designing Systems for Molecular Query-Retrieval and Molecular Informatics. Start date: Feb. 1, 2007. Expires: Jan. 31, 2008. Awarded amount to date: $95,866. Principal investigator: Rahul Singh. Sponsor: San Francisco State University.
Funds a project to improve methods for managing and querying molecular data. The project focuses on three challenges: Designing techniques for representation and similarity-based matching of molecules; development of indexing strategies for molecular query retrieval; and designing knowledge environments for discovery of therapeutics.
Machine Learning Approaches for Genome-wide Biological Network Inference. Start date: May 1, 2007. Expires: April 30, 2009. Awarded amount to date: $345,662.
Principal investigator: Xue-Wen Chen. Sponsor: University of Kansas Center for Research.
Funds an effort to develop computational approaches for uncovering genome-wide networks of interactions between genes and proteins. Research directions include simultaneously integrating multiple biological knowledge into dynamic Bayesian networks for learning networks of gene interactions; learning networks of protein interactions from heterogeneous data; learning integrated networks of gene and protein interactions; learning genome-wide networks of gene and protein interactions; and cross-species network learning.
Algorithms for Experimental Structural Biology. Start date: Feb. 1, 2007. Expires: Jan. 31, 2008. Awarded amount to date: $102,674. Principal investigator: Ramgopal Mettu. Sponsor: University of Massachusetts Amherst.
Supports the development of high-throughput computational methods that can process and interpret experimental data recorded on proteins from X-ray crystallography or nuclear magnetic resonance and synthesize it with other empirically available data, such as protein structures in the Protein Data Bank. Currently, structural biologists “make use of manual or ad hoc approaches for interpreting and synthesizing data that lack guarantees on solution quality or running time (or both),” according to the grantees. Project goals include designing algorithms that efficiently compute optimal or near-optimal solutions, and provide worst-case analysis on solution quality and running time.