MULTIVARIATE ANALYSIS OF STRUCTURE-RELATED DATA TO EXPLAIN MILK CLOTTING ACTIVITY OF PROTEOLYTIC ENZYMES

Publisher Website
Google Scholar
Abstract
Multivariate analysis was used for investigating the relationships between milk-clotting activity and physical/chemical properties, such as CD spectra, hydrophobicity and zeta potential, of ten proteolytic enzymes measured at six different pH values. Cluster analysis of CD data and zeta potential values classified the enzymes into five groups, i.e., three active enzyme groups and two inactive enzyme groups with respect to milk clotting activity. Classification of the enzymes into three groups, i.e., enzymes with high, medium and low milk-clotting activity, was achieved by discriminant analysis after adding the ratio of secondary structure parameters to the predictor variables. Results indicated that β-sheet, β-turn and random structure features were important for milk-clotting activity.