Molecular simulations and machine learning methods for the identification of novel aurora A kinase inhibitors

Aurora A kinase (AAK) is a serine/threonine kinase that stands out as a crucial regulator of mitosis, the complex process of cell division. Notably, the protein AAK plays vital roles in cell cycle regulation and encompasses centrosome maturation, spindle assembly, and chromosome segregation. All such functionalities are essential for ensuring accurate daughter cell formation. Deregulation of AAK expression and activity has been linked to various human diseases, particularly cancer. However, AAK's significance extends beyond normal cellular function. Increased expression or activity of AAK has been implicated in the development and progression of several human cancers. AAK's critical role in cell division and its association with cancer make it a prominent drug target. Herein, series of advance computational approaches was utilized including multi-step molecular docking through AutoDock Vina and PLANTS docking to screen ChemDiv kinase-specific inhibitor library against AAK. Absolute binding energy was estimated, and finally, a molecular dynamics simulation study was conducted to screen out three hit compounds. Both docking studies revealed perfect binding of all identified ligands in active site pockets of AAK protein with similar amino acids of active sites as compared with standard BindingDB_50433632 compound and co-crystal ligand VX-680 binding mode of AAK protein. Therefore, it can be concluded that computational drug discovery approaches are meticulously implemented to identify potential AAKs inhibitors/modulators, and credential of the work was substantiated through the identification of three potential AAKs inhibitors/modulators that may hold significant promise for improving cancer management, however, need extensive biological assays or pre-clinical trials for assessing the efficacy profile of the identified compounds.