Discovery of EGFR kinase's T790M variant inhibitors through molecular dynamics simulations, PCA-based dimension reduction, and hierarchical clustering

Deregulation of epidermal growth factor receptors is one of the major causes of lung cancers, and its kinase has been targeted in associated therapy. Often, mutations causing resistance to therapy have been observed in the patients leading to their poor survival. Even after designing the mutant-specific irreversible inhibitors, the issue to deal with resistance to therapy still remains. Molecular dynamics simulations of biological macromolecules such as proteins offer a greater understanding of biological mechanisms at an atomistic detail. In the present study, principal component analysis-multiple dimension reduction technique and hierarchical clustering was employed in EGFR kinase's T790M mutant form's simulated trajectory to select five diverse conformers for ensemble based virtual screening of 52,000 drug-like molecules from Maybridge library. Further, 50 ns molecular dynamics simulations was conducted on the complexes of top 8 molecules with mutated kinase to validate the binding consistency of these molecules. The strong binding energies and stable dynamics of the simulated complex suggest these molecules to be valuable for drug discovery.