Optimization of enzymatic processing for antioxidant peptides from oyster based on BP neural network

In order to achieve the controllable preparation of antioxidant peptides from oyster,enzymatic hydrolysis conditions were optimized on the basis of the response surface methodology( RSM) and neural network models.In this study,pancreatin was used for the oyster hydrolysis. The results of RSM showed that the optimal hydrolysis condition was 0.28% pancreatin dosage,temperature of 55 ℃,and hydrolysis duration of 3.4 h,under this condition the ·OH clearance rate of its enzymatic prodcut reached 78.4%.The free glutamic acid concentration was used as the monitor of enzymatic hydrolysis process.The back propagation( BP) neural network model was applied to fit for the relationship between enzymatic hydrolysis process and activity of the hydrolysate. The correlation coefficient between predict value and confirmatory experiment were 0.9943,and the average relative error was 2.1%,which indicated that this BP neural network model showed good performance. It was an effective way to monitor the enzymatic hydrolysis on- line,and made it possible to realize controllable preparation of antioxidant peptides.