APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR MODELLING PEA PROTEIN HYDROLYSIS BY TRYPSIN.
 
More details
Hide details
 
Pol. J. Food Nutr. Sci. 2004;54(2):163–168
KEYWORDS
ABSTRACT
The experimentally - determined degrees of hydrolysis (DH) for different enzyme/substrate (E/S) ratios were used for training the artificial neural network (ANN) to predict the trypsin hydrolysis reaction of pea protein for E/S = 15, 25 and 35 mAU/g based on measurements carried out for E/S = 5 mAU/g only. The analysis was performed on a personal computer using a Statistica Neural Networks v. 6.0 PL (StatSoft, Tulsa, OK, USA). The input data were randomized into three sets: learning, validating and testing. Good agreement between the calculated DH values with the use of the ANN designed and the experimentally obtained values was achieved with all three sets of data. The correlation coefficient R2 ranged from 0.978 to 0.999. The differences between the experimental and the calculated data of DH% were very low and ranged from 0.07% to 0.26%. The results obtained show that the ANN method can be useful for the design of the enzymatic processes.
ISSN:1230-0322