Optimizing Bacterial Co-Cultures Using Metabolomics Data
Gregory Medlock of the University of Virginia in the U.S. will develop a method to predict the optimal combinations of different strains of human gut microbes with health-promoting (probiotic) properties to maximize their yield by fermentation and minimize production costs. Microbes tend to grow better when they are in a mixed population (co-culture) because they can share resources and are more resistant to pathogens. Co-culturing can also lower production costs. However, identifying the right mix is challenging as it depends on the properties of each strain and how they interact with other strains. To simplify this, they will build an algorithm that can be applied to any strain of interest. Metabolic and growth profiles will be collected from 10 probiotic strains grown under different conditions to determine their nutrient preferences. These profiles will be used to model the metabolic space occupied by each strain for identifying the combinations that maximize the potential for cross-feeding and minimize resource competition when grown under specified culture conditions. They will test their predictions by comparing biomass produced using the predicted combinations with random combinations and with strains grown on their own.