A network of molecular interactions, involving many thousands of genes, their products, and other molecules, underlie cellular processes. Investigation of these interactions across a wide range of scales ranging from the formation/activation of transcriptional complexes, to the availability of a signaling pathway, all the way to macroscopic processes, such as cell adhesion, calls for a new level of sophistication in the design of genomewide computational approaches. A homogeneous environment for the comprehensive mapping and analysis of molecular cellular interactions in would be a powerful resource for the biomedical research community. We propose the creation of a National Center for the Multiscale Analysis of Genomic and Cellular Networks (MAGNet). The Center will provide an integrative computational framework to organize molecular interactions in the cell into manageable context-dependent components and will develop interoperable computational models and tools that can leverage such a map of cellular interactions to elucidate important biological processes. Center activities will involve a significant, multidisciplinary effort of biological and computational sciences. Specific areas of expertise include natural language parsing (NLP), machine learning (ML), software systems and engineering, databases, computational structural biology, reverse engineering of genetic networks, biomedical literature datamining, and biomedical ontologies, among others. The Center will 1) construct an evidence integration framework to collect and fuse a variety of diverse cellular interaction clues based on their statistical relevance 2) assemble a comprehensive set of physics- and knowledge-based methodologies to fill this framework 3) provide a set of methodologies and filters, anchored in formal domain ontologies, to associated specific interactions to an organism, tissue, molecular, and cellular context. All relevant tools will be made accessible to the biomedical research community through a common, extensible, and interoperable software platform, geWorkbench. We will reach out to train and encourage researchers to use and/or develop new modules for, geWorkbench. An important element of the software platform will be the development of specific components that can exploit the evidence integration techniques developed by Core 1 investigators to combine molecular interaction clues from Core 2 algorithms and databases. Development will be both driven and tested by the biomedical community to ensure the usefulness of the tools and the usability of the graphical user interfaces to address biomedical problems in completely novel ways, to dissect the web of cellular interactions responsible for cellular processes and functions.
Showing the most recent 10 out of 258 publications
