Computer-Assisted Strain Construction And Development Engineering (CASCADE)

This proposal is to respond Army CBD SBIR Phase II invitation for “Computer-Assisted Strain Construction and Development Engineering (CASCADE)”. Our approach is to apply a machine learning system together with publicly available data and private collaboration data for discovering the predictive relations between genomic fingerprints of an organism and its metabolic capabilities, and then use the relations to search for the best organisms to produce desired recombinant proteins or to be used in the applications utilizing the organisms’ metabolic novelty. In Phase II, we will refine and optimize the CASCADE approach from Phase I for the detection, selection, and engineering of such organisms of high value. Specifically, we will integrate the whole CASCADE process into a single screening service of easy-of-use. We will also enhance the system by incorporating more data from publicly available sources. We will collaborate with US Army labs and scientists to work on a couple of novel applications to test and validate the selected organisms through designed experiments. The enhanced genomic information, along with experimental and logic data will be fed into the integrated CASCADE screen service, and the results will be tested in the laboratories to see how well the computer models can assist strain selection, construction and development in real life.