Hybrid of ant colony optimization and flux variability analysis for improving metabolites production

Abstract

Metabolic engineering has been successfully used for the production of a variety of useful compounds such as L-phenylalanine and biohydrogen that received high demand on food, pharmaceutical, fossil fuels, and energy industries. Reaction deletion is one of the strategies of in silico metabolic engineering that can alter the metabolism of microbial cells with the objective to get the desired phenotypes. However, due to the size and complexity of metabolic networks, it is difficult to determine the near-optimal set of reactions to be knocked out. The complexity of the metabolic network is also caused by the presence of competing pathway that may interrupt the high production of a desireable metabolite. Consequently, this factor leads to low Biomass-Product Coupled Yield (BPCY), production rate and growth rate. Other than that, inefficiency of existing algorithms in modelling high growth rate and production rate is another problem that should be handled and solved. Therefore, this research proposed a hybrid algorithm comprising Ant Colony Optimization and Flux Variability Analysis (ACOFVA) to identify the best reaction combination to be knocked out to improve the production of desired metabolites in microorganisms. Based on the experimental results, ACOFVA shows an increase in terms of BPCY and production rate of L-Phenylalanine in Yeast and biohydrogen in Cyanobacteria, while maintaining the optimal growth rate for the target organism. Besides, suggested reactions to be knocked out for improving the production yield of L-Phenylalanine and biohydrogen have been identified and validated through the biological database. The algorithm also shows a good performance with better production rate and BPCY of L-Phenylalanine and biohydrogen than existing results.