Abstract
In this manuscript, three-dimensional quantitative structure-activity relationship (3D-QSAR) studies using comparative molecule field analysis (CoMFA) techniques were applied to provide the structural information of Bcl-2 inhibitors. The CoMFA model obtained from the training set were all statistically significant with the cross-validated coefficients (q2) of 0.568 and conventional coefficients (r2) of 0.991. The contribution of steric field and the electrostatic field is 0.635 and 0.365, respectively. The 3D-QSAR model was furthermore validated by a test set of 6 molecules. The predicted correlation coefficient (r2 pred) on the test set is 0.582. Therefore, the 3D-QSAR models built may be used to exhibit the necessary ligand-based structural environment as well as to design novel Bcl-2 inhibitors with increasing activities.
Keywords: CoMFA, Bcl-2 protein, inhibitors, 3D-QSAR, Comparative Molecular Field Analysis, GPCRs, Sketch Molecule, Gasteiger-Huckel, Partial Least Square, r2pred, HMLP, electrostatic field, leave-one-out, Tripos force fieldCoMFA, Bcl-2 protein, inhibitors, 3D-QSAR, Comparative Molecular Field Analysis, GPCRs, Sketch Molecule, Gasteiger-Huckel, Partial Least Square, r2pred, HMLP, electrostatic field, leave-one-out, Tripos force field