Virtual screening and QSAR modeling for antioxidant activity of benzofurans and flavonoids

Abstract

Drug discovery takes many years and requires big budgets for research and development. QSAR in addition to drug likeness studies contribute strongly to predict and discover new active molecules. Nowadays, the antioxidants are among the most studied and used molecules in drugs and food industries because of their anti-ageing effects. They scavenge free radicals causing oxidative stress. In this work, we perform a QSAR modeling of the antioxidant activity for two sets of benzofurans and flavonoids by artificial neural networks and Gaussian process seldom used in this approach. Their predictability coefficient was acceptable with a value that exceeds 0.6. Drug likeness studies based on Lipinski and Veber rules, besides the lipophilicity indices permitted to define the drug like molecules.