A Three-Dimensional Quanititative Structure-Activity Relationship (3D-QSAR) Model for Predicting the Enantioselectivity of Candida antarctica Lipase B

Computational techniques involving molecular modeling coupled with multivariate statistical analysis were used to evaluate and predict quantitatively the enantioselectivity of lipase B from Candida antarctica (CALB). In order to allow the mathematical and statistical processing of the experimental data largely available in the literature (namely enantiomeric ratio E), a novel class of GRID-based molecular descriptors was developed (differential molecular interaction fields or DMIFs). These descriptors proved to be efficient in providing the structural information needed for computing the regression model. Multivariate statistical methods based on PLS (partial least square - projection to latent structures), were used for the analysis of data available from the literature and for the construction of the first three-dimensional quanititative structure-activity relationship, (3D-QSAR) model able to predict the enantioselectivity of CALB. Our results indicate that the model is statistically robust and predictive.