Computer-Aided Drug Design Targeting Protein Phosphorylation

University Of Missouri-St. Louis, Saint Louis MO

Investigators

Linked publications, trials & patents

Abstract

The long-term goal of this research is the development and application of computational methods to aid the design of drugs targeting protein kinases and related proteins. The approval of about 100 protein kinase inhibitors as drugs for cancer treatment and other ailments has demonstrated protein kinases as important drug targets. With over 500 protein kinases present in humans and with many mutants driving diseases, many more drugs can be developed by targeting protein kinases. This renewal application focuses on improving and validating simulation methods for finding compounds with therapeutically useful drug-binding kinetics, and in using molecular simulation and docking methods to help find drugs to treat lung cancer patients carrying specific mutations in the epidermal growth factor receptor (EGFR). Specific Aim 1a continues to improve the milestoning method for simulating drug-binding kinetics by using better progress coordinates. Specific Aim 1b performs systematic sensitivity analysis to identify key amino acids and molecular interactions in protein kinase-ligand systems that determine their dissociation kinetics. Specific Aim 2 will help predict drug response of lung cancer patients carrying mutations in EGFR. It contains two sub-aims. Specific Aim 2a performs molecular simulations on mutants of EGFR found in lung cancer patients to examine their likelihood of driving disease. Specific Aim 2b extends a machine-learning enhanced ensemble docking method to predicting drug binding to disease-driving mutants of EGFR. The research is significant in moving simulation methods closer to the stage of practical applications in proposing compounds that are more likely to present therapeutically useful drug-binding kinetics, and in developing predictive models to help decide which mutants of EGFR found in lung cancer patients can benefit from existing drugs. It is novel in extending systematic sensitivity analysis to identify key amino acids and molecular interactions in a protein kinase-ligand system in determining drug-dissociation kinetics, information useful for deciphering the mechanism of drug-dissociation kinetics and designing drugs with therapeutically useful drug-binding kinetics. The predictive model for drug response is unique in considering both the possibility of a mutant of EGFR to drive disease and to bind drugs. The prediction of drug binding to EGFR mutants by ensemble docking is enhanced by machine learning leveraging experimental data of drug response for mutants found in lung cancer patients. Publications and preliminary results have demonstrated the feasibility of the approaches.