Algorithm and software for modelling of food protein hydrolysis kinetic

The objective of this research was to propose two numerical approaches and based on these approaches user-friendly dialogue software for determining the kinetics of enzyme hydrolysis of protein at different initial reaction parameters, which could be an initial substrate and enzyme concentrations, temperature and pH-value. The proposed approaches mainly are based on the utilization of adaptive random search technique, Artificial Neural Networks (ANNs) and the following well-known exponential kinetic equation presented a relationship between reaction rate and enzymatic reaction parameters dh/dt = aexp(-bh), where coefficients a and b have different expressions according to different reaction mechanism. The first approach consists of using the adaptive random search technique for fitting the kinetic constants of enzymatic reactions to a set of given experimental or time course profiles (TCPs) in the form of degree of hydrolysis over reaction time. Use of this technique can give much greater confidence that if a global optimal solution had existed within the global domain then it would not have been missed by the search routine. The second approach uses ANNs for estimating the values of coefficients a and b of the exponential kinetic equation. In this case each of experimental TCPs is fitted to this equation by using adaptive random search technique. Optimal a and b values obtained for TCPs and initial enzymatic reaction parameters are used as an input data for the ANNs learning process. The user-friendly dialogue software “ANNEKs” (ANN Enzyme Kinetics) realized two mentioned above approaches is developed. Findings from the work reported in this study would suggest that the developed user-friendly interfaces and utilized numerical approaches make the “ANNEKs” software package useful for food scientists and engineers.