Dell Technologies HPC Community Event
About the Event
Drug discovery is an expensive process that takes an average of $3 billion dollars and about 10 years to bring a new drug from the lab bench to a patient. Two of the main challenges are 1) low hit discovery rate of conventional high-throughput screening (HTS) experimental approaches and 2) unwanted side-effects of small molecules which tend to cross-react with unintended targets. While virtual docking can help to address these challenges and can significantly accelerate the overall process of drug discovery, it ultimately lags the current rate of expansion of chemical databases that already exceed billions of molecular records. A recent surge of small molecule availability presents great drug discovery opportunities but also demands much faster screening protocols. In order to address this challenge, we herein introduce Deep Docking (DD), a novel deep learning platform that is suitable for docking billions of molecular structures in a rapid yet accurate fashion.
The DD approach utilizes DNN models trained on docking scores of subsets of a chemical library to approximate the docking outcome for yet unprocessed entries and, therefore, to remove unfavorable molecules in an iterative manner. Such modeling approach, implemented in an iterative manner on GPU platform enabled up to 100 folds speed up of docking proses, without notable loss of information (retaining >90% of qualified candidate compounds).
Thus, the DD methodology, enables drug discoverers to access ultra-large chemical libraries, consisting of hundreds of billions of structures, to screen them for potential drug candidates. We have recently applied DD to screen the Enamine REAL Space library of 40B compounds with 5 independent docking programs and the SARS-CoV-2 3CL enzyme as a drug target. The resulting consensus predictions based on 200B generated docking scores lay a foundation for the identification of much-needed effective therapeutics for COVID-19.
About the Speaker
Artem Cherkasov, Professor, Faculty of Medicine, University of British Columbia
Artem Cherkasov is a Professor of Medicine at the University of British Columbia (Vancouver, Canada) and a Director of Therapeutics Development at Vancouver Prostate Centre. Research interests include computer-aided drug discovery (CADD), A.I., QSAR modeling, drug reprofiling, and the development of new cancer and COVID therapies.
Dr. Cherkasov co-authored more than 200 research papers, 80 patent filings, and several book chapters. During his tenure at the UBC, Dr. Cherkasov has been a principal applicant or co-applicant on a number of successful grants totaling over 80M dollars and licensed 8 drug candidates to big pharma companies, major international venture funds, and spin-off companies