Quantitative structure activity relationship (QSAR) study of antioxidant activity of catechin derivative

ABSTRACT: QSAR is a computational method used in identifying the relationship between the structure of a compound and its biologival activities. In this study, quantitative-structure activity relation model for antioxidant activity of twenty-three catechin derivatives was generated. Both the structure of compounds and its biological activity was obtained from literature. The data set was divided into training set (n=17) and test set (n=6). The compounds were sketched in 2D before the computation of the physicochemical and topological descriptors. QSAR modelling was done using multiple linear regression method (MLR). The predictive power of these models were evaluated internally by applying the following statistical parameters for the training set and test set: RMSE (root-mean-square error for prediction), squared correlation coefficient (R2), adjested squared correlation coefficient (R2 adj). difference between R2 and R2 adj (R2 – R2 adj), squared cross-validation correlation coefficient coefficient (Q2). External validation was also performed by employing Golbraikh and Tropsha criteria. The model produced showed the three descriptors with highest influence on the antioxidant activity of the compounds were all under topological descriptions. The two connectivity index descriptors showed a negative contribution on the antioxidant activity of the compound while the topological charge index descriptor has a positive contribution. Both the internal (RMSE=0.22000; R2=.87713; R2 adj=0.84877; R2-R2adj=0.02836; Q2=0.81994) and external validation (R2=0.85956; Ro2=0.85956; R2-R02=0.03524; R2=0.03524; R2-R02/R2=0.04234; k=0.99266; Q2=0.81994) of the model showed that the model obtained from MLR method is acceptable and can be used to predict antioxidant activities of catechin compounds.

5

Filipiniana

5