Indian jujube (Ziziphus mauritiana L.) fruit juice extraction using cellulase enzyme: Modelling and optimization of approach by ANN-GA

Abstract

The Indian jujube is a seasonal fruit that is beared in abundant during February and March in India. The fruit is rich in ascorbic acid and bioactive compounds like total phenolic compounds and antioxidant activity. It is a significant source of revenue for the local tribes. However, due to a lack of scientific documentation and value addition, a substantial amount of fruit is thrown away. As a result, making juice from Indian Jujube pulp will be a cost-effective and straightforward option that will extend its shelf life and aid in its commercialization. A Box–Behnken design was generated with 17 experimental runs was modelled using an artificial neural network by taking independent variables as incubation temperatures (30, 40, and 50 °C), incubation duration (40, 70, and 100 mins) and enzyme concentrations (0.01, 0.055, and 0.1 uM) while taking dependent variables as yield, viscosity, refractive index, total soluble solids, polyphenol, antioxidant activity, and protein content. The model was fitted successfully with a maximum R-value of 0.9997 and a least MSE 0.0978. The weight and bias values obtained by artificial neural network modelling was used by a genetic algorithm to estimate the fitness function, which minimizes dependent variables like viscosity and maximizes other dependent variables. The optimized independent variables obtained were 30 °C incubation temperature, 58 min of incubation duration and 0.1 uM of enzyme concentration with mean and best fitness values of -187.238 and -187.304, respectively. The optimized independent variables were validated by physiochemical experiment, which resulted in juice yield (80.89%), viscosity (3.11 cP), refractive index (1.3549), total soluble solids (14.20 °Brix), polyphenolic content (0.94 g GAE/100 mL), antioxidant activity (82.41%) and protein content (9.53 mg/100 ml). The results revealed that the observed effects of the responses of enzyme-treated were significantly higher than those of the juice without treatment. Therefore, these conclusions highlight the highly effective method of juice extraction with a better statistical approach.