The ultimate goal of this research is to integrate visual and computational descriptions of complex metabolic and regulatory networks to aid biologists in evaluating hypotheses for how these dynamic networks function under different conditions. These networks will combine graph models from pathway databases, text-mining programs, machine learning systems, and other sources, together with multiple classes of experimental data. The unique features of the proposed graph visualization and analysis platform are: 1) Evaluation of the structural effects of dynamic links that change depending on time and other conditions. 2) The ability to immediately integrate current research hypotheses with available published results to evaluate their impact and explanatory power. 3) Interactive display of large metabolic and regulatory networks in either user- or automatically selected levels of detail. 4) Creation of visual graph display tools specifically designed for improved biological network display, comparison, and analysis. As part of this process, significant problems in the analysis of variable graph structures, incremental graph layout, and effective visualization and labeling will be addressed with an interdisciplinary focus. The software will be open source and freely available to academic institutions. It will be evaluated and validated using three interrelated but only partially understood signal transduction networks: ethylene, jasmonate and salicylic acid. These pathways interact complexly to direct specific plant defense responses to stress. The software will contribute to the research community's open source software toolbox by providing more effective ways to visualize and manipulate graphs. By actively integrating biologists in the design and development we will ensure practical applicability and usability for non-computer-expert users. Education and outreach activities will promote research, K-12 and undergraduate education, and dissemination of results to a broad audience, while developing a new generation of scientists that employ the powers of computers to their fullest to advance all sciences.
