Optimization of enzymatic processing for α-Glucosidase inhibitor from scallop skirts on BP neural network

In order to obtain the active hydrolysate with good α-glucosidase inhibition activities,scallop skirts were used as the raw material which were hydrolyzed by the selected compound protease. To simulate and predict these hydrolysis processes,a Back-Propagation neural network was established. Parameters of the enzyme quantity,the substrate concentration and the hydrolysis duration were used as inputs,and the corresponding output was inhibitory rate of α-glucosidase. The optimization of enzymatic hydrolysis process was carried out by the genetic algorithm method. Results showed that the maximum inhibitory rate of the hydrolysate could reach 95.5%±2.6% under conditions of enzyme quantity 140 U/mL,substrate concentration 2%,hydrolysis duration 4.4 h,respectively. In this study,the established model could accurately predict and monitor the enzymatic hydrolysis of scallop skirts. This study could provide basis data for further bioactivity researches and utilizations of scallop skirts.