NECom

NECom predicts Nash equilibria in microbial metabolic interactions by modeling continuous metabolic-flux strategy spaces with a bi-level optimization framework to analyze evolutionary game-theoretic outcomes in microbial communities.

Key Features:

• Bi-level Optimization Framework: Employs a bi-level optimization framework inspired by evolutionary game theory that ensures feasible solutions are Nash equilibria of microbial community metabolic models regardless of an outer-level community objective.
• Continuous Strategy Space Modeling: Models continuous interdependent strategy spaces of metabolic fluxes rather than relying on discrete matrix-game representations.
• Prevention of 'Forced Altruism': Formulation prevents 'forced altruism' by enabling sensing of potential metabolite exchanges to identify evolutionarily favorable interactions between community members.
• Predictive Accuracy: Demonstrates improved prediction of classical game outcomes such as prisoner's dilemma, snowdrift, and cooperation in metabolic contexts compared to existing methods.

Scientific Applications:

• Insights into Mutualism: Applied to an algae–yeast co-culture with cross-feeding features related to lichen, providing insight into why mutualism can be evolutionarily favorable despite the costs of cross-feeding metabolites.
• Comparison with Matrix Games: Highlights similarities and differences between games in continuous metabolic flux space and traditional matrix games to deepen understanding of microbial interactions.

Methodology:

Uses a bi-level optimization framework and continuous metabolic-flux strategy modeling; simulates various growth conditions without training parameters on experimental data and was evaluated across 488 growth conditions corresponding to 3,221 experimental data points, with predicted species growth rates compared to flux balance analysis yielding root-mean-square error reductions of 63.5% and 81.7% for the two species studied.