QSAR studies on 4-anilino-3-quinolinecarbonitriles as Src kinase inhibitors using robust PCA and both linear and nonlinear models

Abstract

Quantitative structure-activity relationship (QSAR) studies have been carried out on 4-anilino-3-quinolinecarbonitriles, a set of novel Src kinase inhibitors, with the aim of dissecting the structural requirements for Src inhibitory activities. After outlier identification using robust principal component analysis (robust PCA), linear models based on forward selection combined with multiple linear regression, (FS-MLR), enhanced replacement method followed by multiple linear regression (ERM) and a nonlinear model using support vector regression (SVR) were constructed and compared. All models were rigorously validated using leave-one-out cross-validation (LOOCV), 5-fold cross-validations (5-CV) and shuffling external validation (SEVs). ERM seems to outperform both FS-MLR and SVR evidenced by better prediction performance (n = 35, R²_training = 0.918, R²_pred = 0.928). Robustness and predictive ability of ERM model were also evaluated. The generated QASR model revealed that the Src inhibitory activity of 4-anilino-3-quinolinecarbonitriles could be associated with the size of substituents in the C7 position and the steric hindrance effect. The results of the present study may be of great help in designing novel 4-anilino-3-quinolinecarbonitriles with more potent Src kinase inhibitory activity.

Keywords::

• QSAR
• robust PCA
• ERM
• 4-anilino-3-quinolinecarbonitriles
• Src kinase inhibitors

Acknowledgment

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.