Identification of Aurora A kinase allosteric inhibitors: A comprehensive virtual screening through fingerprint-based similarity search, molecular docking, machine learning and molecular dynamics simulation

Abstract

The Aurora A kinase (AAK) protein controls spindle assembly and promotes cell divisions in various diseases including cancer. In the present study, allosteric inhibition of AAK protein through different advanced computational screening approaches is employed to target AAK’s allosteric inhibition-modulation. Precisely, extensive computational techniques including allosteric binding sites recognition of AAK protein, fingerprint-based similarity search, multi-step molecular docking through AutoDock Vina and PLANTS, and a 100 ns molecular dynamics (MD) simulations studies were carried out followed by the calculation of binding free energy with MM-GBSA based approaches, has been employed for identification of potential allosteric inhibitors-modulators of AAK protein. The study outcome highlighted that all three identified small molecular chemical entities exhibit strong binding interaction affinity in both the docking analyses at the allosteric domain of AAK protein and also greater interaction stability in comparison to the considered standard compound. In addition, all identified three screened hits also show acceptable pharmacokinetics and medicinal chemistry properties, which certainly dictates their potentiality for becoming good drug-like compounds for inhibiting-modulating the activity of AAK protein binds with similar amino acids of the allosteric domain of AAK protein with selected three compounds. All the selected molecules were found to show an acceptable ADMET profile. Moreover, the MM-GBSA-based binding energy was found to be in the range of −8 to −35 Kcal/mol, which showed a strong association between proposed molecules and AAK protein. Comprehensive computational approach shows that the selected proposed three inhibitors of AAK protein are the best candidates as potential inhibitors.