A hybrid of differential search algorithm and flux balance analysis to: Identify knockout strategies for in silico optimization of metabolites production

Abstract

An increasing demand of naturally producing metabolites has gained the attention of researchers to develop better algorithms for predicting the effects of reaction knockouts. With the success of genome sequencing, in silico metabolic engineering has aided the researchers in modifying the genome-scale metabolic network. However, the complexities of the metabolic networks, have led to difficulty in obtaining a set of knockout reactions, which eventually lead to increase in computational time. Hence, many computational algorithms have been developed. Nevertheless, most of these algorithms are hindered by the solution being trapped in the local optima. In this paper, we proposed a hybrid of Differential Search Algorithm (DSA) and Flux Balance Analysis (FBA), to identify knockout reactions for enhancing the production of desired metabolites. Two organisms namely Escherichia coli and Zymomonas mobilis were tested by targeting the production rate of succinic acid, acetic acid, and ethanol. From this experiment, we obtained the list of knockout reactions and production rate. The results show that our proposed hybrid algorithm is capable of identifying knockout reactions with above 70% of production rate from the wild-type.