PREDICTING PROTEIN FUNCTIONALITY WITH ARTIFICIAL NEURAL NETWORKS - FOAMING AND EMULSIFYING PROPERTIES

Using physicochemical properties of 11 food-related proteins, artifici al neural networks (ANN) were developed for predicting foam capacity a nd stability and the emulsion activity index. The prediction accuracy of ANN was compared to that of principal component regression (PCR) mo dels. ANN had better prediction ability than PCR, especially after cro ss-validation. For foam stability, PCR did not generate a significant model. ANN and PCR models indicated that fluorescence probe hydrophobi city (measured using cis-parinaric acid) and other properties, such as viscosity, surface tension and net charge were important in predictin g foam capacity and stability.