Statistical and evolutionary optimisation of operating conditions for enhanced production of fungal l-asparaginase

Abstract

AbstractA three-level central composite design of the Response Surface Methodology and the Artificial Neural Network-linked Genetic Algorithm were applied to find the optimum operating conditions for enhanced production of l-asparaginase by the submerged fermentation of Aspergillus terreus MTCC 1782. The various effects of the operating conditions, including temperature, pH, inoculum concentration, agitation rate, and fermentation time on the experimental production of l-asparaginase, were fitted to a second-order polynomial model and non-linear models using Response Surface Methodology and the Artificial Neural Network, respectively. The Artificial Neural Network model fitted well, achieving a higher coefficient of determination (R 2 = 0.999) than the second-order polynomial model (R 2 = 0.962). The l-asparaginase activity (38.57 IU s mL−1) predicted under the optimum conditions of 32.08°C, pH of 5.85, inoculum concentration of 1 vol. %, agitation rate of 123.5 min−1, and fermentation time of 55.1 h was obtained using the Artificial Neural Networklinked Genetic Algorithm in very close agreement with the activity of 37.84 IU mL−1 achieved in confirmation experiments.