An improved cellular automata model of enzyme kinetics based on genetic algorithm

highlights
• Cellular automata are used to model enzymatic reaction kinetics on a 2D torus.
• Fine-tuning the probabilities of reactions and movement is challenging.
• A multi-objective genetic algorithm obtains the probabilities in feasible time.
• A validation study indicates the applicability of the cellular automata model.

Based on a generalized Cellular Automata (CA) model which has the flexibility of reaction interactions between all possible two-agent combinations on both reactant and product side, in previous studies by the authors, probabilistic reaction events were combined in a multiplicative order and an exploratory study of the variability analysis was performed. However it was found that the probabilistic parameter values are extremely sensitive to the final model output. As such, NSGA-II based on a Multi-objective Genetic Algorithm (MOGA) is used in this study to obtain optimum sets of these probability parameter values within a feasible computational time. For each generation of the MOGA, the probability rules parameter values are updated (improved) based on the objective functions. Finally, a validation study is performed to indicate the applicability of the CA model to represent specific enzymatic reactions when coupled with the multi-objective optimization algorithm.