Improved Metaheuristic Algorithms for Metabolic Network Optimization

Joint Conference on Green Engineering Technology & Applied Computing 2019

工业技术（总论）;计算机科学

Mohd Daud, K.^1 ; Zakaria, Z.^1 ; Hassan, R.^2 ; Mohamad, M.S.^3^4 ; Ali Shah, Z.^1

Artificial Intelligence and Bioinformatics Research Group, School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia, Johor Bahru

81310, Malaysia^1

Software Engineering Department, School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia, Johor Bahru

81310, Malaysia^2

Institute for Artificial Intelligence and Big Data, Universiti Malaysia Kelantan, City Campus, Pengkalan Chepa, Kota Bharu, Kelantan

16100, Malaysia^3

Faculty of Bioengineering and Technology, Universiti Malaysia Kelantan, Jeli Campus, Lock Bag 100, Jeli, Kelantan

17600, Malaysia^4

关键词: Artificial bee colonies (ABC);    Comparative performance;    Comparative studies;    Computational time;    Flux balance analysis;    Meta heuristic algorithm;    Meta-heuristics algorithms;    Optimization problems;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/551/1/012065/pdf
DOI  :  10.1088/1757-899X/551/1/012065

来源: IOP

【 摘 要 】

Metaheuristic algorithms have been used in various domains to solve the optimization problem. In metabolic engineering, the problem of identifying near-optimal reactions knockout that can optimize the production rate of desired metabolites are hindered by the complexity of the metabolic networks. Through Flux Balance Analysis, different metaheuristics algorithms have been improved to optimize the desired phenotypes. In this paper, a comparative study of four metaheuristic algorithms have been proposed. Differential Search Algorithm (DSA), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC) and Genetic Algorithm (GA) are considered. These algorithms are tested on succinic acid production in Escherichia coli. The comparative performances are measured based on production rate, growth rate, and computational time. Hence, from the results, the best metaheuristic algorithms to solve the metabolic network optimization is identified.