In silico design of potent antioxidant compounds using 2D-QSAR Molecular docking ADMET prediction and Molecular dynamic simulation

Background Oxidative stress caused by free radicals plays a major role in the development of aging and several degenerative disorders, including cancer, cardiovascular diseases, diabetes, and Alzheimer’s disease. Therefore, the discovery of potent antioxidant agents remains an important research focus. This study aimed to design new antioxidant compounds (benzoxazine derivatives) using a ligand-based approach involving QSAR analysis, molecular docking, molecular dynamic simulation, and ADMET prediction. Methods A QSAR model was developed using twenty-eight benzoxazine derivatives with the aid of Material Studio v8.0. Applicability domain and virtual screening were employed to identify a suitable template compound, which was subsequently modified to design new potent analogues. Molecular docking was performed against cytochrome c peroxidase (PDB ID: 2XO8) using Molegro Virtual Docker software. SwissADME and pkCSM servers were used to predict drug-likeness and pharmacokinetic properties. Results Among the four generated QSAR models, model four was selected as the best model based on its statistical parameters (R ² _ext  = 0.7449, R ² _trn  = 0.7889, R ² _adj  = 0.7361, Q ² _LOO  = 0.6889, and cR²p \(\:=0.6015\)). The designed compounds exhibited improved predicted antioxidant activities (pIC ₅₀ = \(\:8.002-8.532\)), compared with the template and ascorbic acid (pIC ₅₀ = 7.817). Docking results revealed D5, D1, and D4 as having notable MolkDock scores of \(\:-135.37\) kcal/mol, \(\:-134.\)38 kcal/mol, and \(\:-131.14\) kcal/mol respectively. Furthermore, the molecular dynamic simulation and ADMET results revealed the stability of D1 and D2 design complexes, favorable drug-likeness, and good pharmacokinetic profiles. Conclusions The designed benzoxazine derivatives may serve as promising antioxidant agents subject to further in vitro and in vivo experiments.