Although experimentation is key for the development of novel fermentation processes and organisms to economically produce bioproducts from biomass, significant resources (time and cost) can be saved through computational modeling. Computational models of metabolic, transcriptional regulatory, and signaling networks can predict environmental or genetic parameters which can be manipulated to achieve optimal fermentation performance.
An example metabolic network for central metabolism viewed in the Matlab Simbiology toolbox.
Our research group develops and integrates computational models of Saccharomyces cerevisiae and Thermotoga neapolitana with our experimental results. The experimentation and modeling work is a feedback system; we iteratively refine our models with our experimental data and generate new hypotheses to test in vivo. Using a constraint-based approach to conduct simulations with genome-scale reconstructed networks allows us to predict phenotypes from genetic modifications, and allows us to predict an organism's response to changes in fermentation composition.
The Walker group has the capability to build reconstructed networks, as well as to apply reconstructed networks published by other research groups to our research efforts. The group conducts a variety of computational tasks, including building and modifying the stoichiometric matrix, simulating growth with flux-balance analysis, performing dynamic flux-balance analysis, simulating growth phenotypes in organisms with combinations of gene deletions, performing flux variability analysis, network robustness analysis, and network module identification, as well as other computational and systems biology approaches.