Development of Predictive Software Tools to Construct and Analyze Dynamical Networks for GTL Systems Biology Knowledgebase

Recent technological advances continue to accelerate the rate of data availability from a number of omics platforms such as genomics, proteomics, metabonomics and metabonomics. This has served as a driving force in efforts to develop novel predictive software tools that can model the underlying biological system by inferring mechanisms involved in generating responses to external stimuli such as environmental effects. However, the lack of in vivo kinetics for a large number of intracellular biochemical reactions limits the predictive capabilities of available tools. In this SBIR, CFDRC will develop an integrated predictive modeling software toolkit to address this limitation enabling fast characterization of the effects of environment on phenotypes. In the Phase I effort, we will develop a prototype using genomic data and extend it to utilize other types of omic data sources of interest in the Phase II. This integrated toolkit will be able to (1) identify significant biological features by comparing omic data from the control and affected organisms; (2) use the identified significant biological features to construct a comprehensive network model of the system incorporating the biological pathways involved; and (3) simulate the constructed network model by means of a Boolean Network Dynamics Target Identification (BNDTI) framework. Results obtained from BNDTI will allow prediction of the effect of changes in environmental factors on the organisms phenotype. The software toolkit will be validated against known examples of applied stimuli and observed phenotypes in model organisms, and the dynamic networks constructed in this effort will be refined as new interactions are found. Commercial Applications and Other Benefits: This software toolkit will provide enhanced model construction and analysis capabilities with applications in a number of areas. In particular, this includes the ability to identify and rank critical nodes by means of a Boolean approach applied to the constructed network, which can lead to development of strategies for targeting the identified critical network nodes. Applications of this framework include discovering new approaches to increase energy yield or to improve the efficacy of an administered drug. The functionality of the software toolkit will be made accessible to the community via web, allowing users them to specify the organism and select omic datasets for analysis, along the guidelines of Systems Biology Knowledgebase framework. The proposed approach is of particular relevance to the DOE community that aims to develop better alternative energy sources or cleanup or toxic waste and other similar goals where it is desirable to find specific strains of microbes that are highly effective. Further, integrating this framework with a data storage and retrieval system will lead to the development of a very efficient Laboratory Information Management System that can be used tasks and access results. Other potential applications include areas such as development of new antibiotics to treat infections and to identify, characterize and effectively treat exposures to toxic compounds.