In silico molecular investigations of derived cyclohexane-1,3-dione compounds as potential inhibitors of protein tyrosine kinase C-met: 2D QSAR, molecular docking and ADMET

Abstract

The C-met receptor tyrosine kinase represents an interesting anti-cancer target. In this work, we present a theoretical study of the quantitative structure-activity relationship, QSAR, inhibitor of the enzymatic activity of said C-met protein. Using statistical techniques, RLM, RNLM and Y-randomization assay of the field of applicability, we studied a series of 36 molecules derived from cyclohexane-1,3-dione, dimedon, as anticancer agents capable of inhibiting C-met receptor tyrosine kinase. In this study we developed models showing excellent statistical results for multiple linear regression $\text{R}^{2}$=0,913; $\text{R}^{2}$ cv=0,85, $\text{R}_{\text{t}\text{e}\text{s}\text{t}}^{2}$=0,934) and for multiple nonlinear regression ($\text{R}^{2}$=0,991$;\text{R}^{2}$ cv=0,82; $\text{R}_{\text{t}\text{e}\text{s}\text{t}}^{2}$ = 0,997). These results demonstrate the great ability of multiple linear regression to effectively model the inhibitory activity of the enzymatic activity of the C-met protein and its predictive capacity. Motivated by these results, we designed 16 molecules adopted for the treatment of non-small cell lung cancer (NSCLC) to evaluate the properties of ADMET in silico which will be supplemented by a molecular Docking.