Comparative QSAR modeling of antitumor activity of ARC-111 analogues using stepwise MLR, PLS, and ANN techniques

摘要

Dibenzo [c,h]-[1,6] naphthyridin-6-one (ARC-111) and its analogues can exhibit potent topoisomerase I targeting and antitumor activities. Quantitative structure-activity relationship (QSAR) studies on the antitumor activity of 22 ARC-111 analogues to RPMI 8402 (expressed as pIC₅₀) were performed by using electronic, spatial, thermodynamic, and topological descriptors based on different chemometric tools including stepwise multiple linear regression (stepwise MLR), partial least squares (PLS), and artificial neural network (ANN). QSAR equations were examined using a training set of 18 compounds and an external test set of four compounds, which was divided by cluster analysis. The ANN model was the most powerful, with a square of predictive correlation coefficient R _pred 2 of 0.963 for the test set; the R _pred 2 values for stepwise MLR and PLS were 0.780 and 0.729, respectively. The results obtained with these models indicated that the antitumor activity of ARC-111 analogues depends strongly on electronic factors (such as sum of E-state of H-bond acceptors, dipole moment as debyes, and total energy).