Exploring Novel Plumbagin Derivatives as Potential Inhibitors For HepG2 Cell Line Using QSAR And Molecular Docking Models

In this study, a series of new plumbagin (PLB) derivatives from Plumbago Zeylanica Linn were discovered as potential inhibitors against liver cancer (HepG2 cell line). The derivatives were newly designed and developed by using quantitative structure-activity relationship (QSAR) models. The QSAR models were built by using multiple linear regression (MLR-QSAR) and artificial neural networks (ANN-QSAR) techniques through the IC50 values and molecule descriptors of experimental PLB derivatives. Furthermore, the derivatives passed rigorous evaluation for drug-likeness using ADME analysis. Docking simulation was applied molecular docking into the VEGFR-2 receptor for the final analysis to find out the good plumbagin derivatives. The successful MLR-QSAR model consists of 6 descriptors, which met the statistical requirements:
$R_{\text{train}}^{2}= 0.964; R_{\text{adj}}^{2}= 0.960$
and
$Q_{\text{CV}-\text{LOO}}^{2}= 0.953$
. The ANN-QSAR model with I(6)-HL(5)-O(I) architecture was found out on the basis of variables of MLR-QSAR model, with statistical data: R
²
train = 0.963;
$Q_{\text{validation}}^{2}= 0.996$
and
$R_{\text{test}}^{2}= 0.993$
. As a result, 19 PLB (PLB1-PLB19) derivatives had good IC
50
values which were calculated by using two models because they were in the applicability domain (AD); in which, 10 derivatives met the criteria of drug-likeness. Finally, molecular docking showed that four novel PLB derivatives are potential derivatives in anti-cancer HepG2 by validating the VEGFR-2 receptor; in which PLB4 is the most potential derivative. The results are of potential value for experimental drug development in the treatment of liver cancer.