Speaker: Peter Coveney, professor of physical chemistry, University College London

Seminar: Rapid Free Energy Prediction for Ligand-Protein Systems — Optimized Calculations for Drug Discovery and Personalized Medicine

Host: Professor Hendrik Heinz 

Abstract
Free energy estimation is a key step in the early stages of drug discovery. It is intricately bound up in the uncovering of molecules which have the potential to reach the market. However, the pharma industry’s record of successful drug discovery is not one to be proud of — it is far too long, far too expensive and far too unsuccessful. Our approach seeks to embrace a more holistic approach, in which ultimately all stages of the pre-clinical workflow can be executed on a sufficiently large computer prior to moving into the laboratory and the clinic.  In my talk, I will first focus on how molecular simulation, particularly molecular dynamics, is being used to provide reliable free energies from early stage hit discovery to lead optimization and even into personalized drug selection. The calculations turn on ensuring that molecular dynamics is applied reliably and reproducibly using ensemble methods so as to yield actionable predictions.

The second part of my talk will describe the construction of a large and complex computational workflow which is designed to recommend a diverse set of molecules as reliable candidates for laboratory based synthesis, testing and development. This is achieved by bringing the aforementioned physics-based methods into interaction with generative active-learning and other forms of artificial intelligence to propose entirely new molecules. The workflow is executed on the world’s largest and fastest supercomputer, Frontier, which resides at the DOE’s Oak Ridge National Laboratory.

Bio
Peter Coveney is a professor of physical chemistry, honorary professor of computer science and director of the Centre for Computational Science (CCS) and associate director of the Advanced Research Computing Centre at University College London (UCL). He is also professor of applied high performance computing at the University of Amsterdam (UvA) and professor adjunct at the Yale School of Medicine, Yale University. He is a fellow of the Royal Academy of Engineering and  member of Academia Europaea. Professor Coveney has made outstanding contributions across a wide range of scientific and engineering fields, including physics, chemistry, chemical engineering, materials, computer science and high performance computing and biomedicine, much of it harnessing the power of supercomputing to conduct original research at unprecedented space and time scales. He has shown influential leadership across these fields, manifested through running multiple initiatives and multi-partner interdisciplinary grants, in the UK, Europe and the US. In addition to his scientific writings and publications, he has published three books for the general reader, The Arrow of Time, Frontiers of Complexity and Virtual You; a fourth, Molecular Dynamics: Probability and Uncertainty, is due in 2025.