Generating a potent inhibitor against MCF7 breast cancer cell through artificial intelligence based virtual screening and molecular docking studies

Abstract

Artificial Intelligence (AI) has been widely adopted by pharmaceutical industry to aid rationally drug design and development by fostering the quick delivery of drug targets with optimized structures in spite of huge chemical space of >1060 drug molecules. Tamoxifen, Selective Estrogen Receptor Modulator (SERM), is the drug for breast cancer cell, MCF 7 with many side effects. Tamoxifen may cause side effects like increased bone or tumor pain, pain or reddening around the tumor site, hot flashes, nausea and excessive tiredness etc., Therefore, compound which can resist ER’s bioactivity is considered as an important target for treating breast cancer. In this study, AI based Virtual Screening (VS) method using an efficient Generative Neural Network (GNN) model has been experimented to generate high inhibitory potential hit drug-like inhibitors. Physicochemical, Pharmacokinetic and toxicity analysis are carried out for conforming the sub-selection of drug-likeness of inhibitors. Additionally, Molecular Docking studies with DNA (355D) and protein (3EU7) are performed for the evaluation of binding affinity, prediction of intermolecular interactions and inhibition constant. The docked results of the inhibitor M22 (methyl 2-[(2-benzoylphenyl) carbamoyl] benzoate) has low free energy of binding (-8.61 Kcal/mol and -8.05 Kcal/mol) and low Inhibition constant, Ki, value (0.486 μM and 1.25 μM) as compared to Tamoxifen (-6.7 Kcal/mol & -5.62 Kcal/mol and 12.2 μM & 75.85 μM). Thus, minimum amount of the M22 inhibitor is enough as compared to Tamoxifen and M22 has 3 benzene rings, extended conjugation, amide linkage and huge number of labile electrons which facilitates as a potent drug. This study provides a greenish path to synthesise a potent inhibitor, M22, for further experimental studies rather than preparing number of inhibitors on the atom economy way.