Grant: Discovery Grants Program - Individual

The emergence of large (>100 million molecules) make-on-demand chemical libraries represents a new paradigm that is already transforming the research landscape of drug discovery. Docking of billions of synthesizable small molecules has demonstrated great potential to identify novel drug-like compounds with unprecedented in vivo and in vitro bioactivity. Unfortunately, conventional virtual screening remains challenging at the level of billions of molecules due to its high computational cost, and the ever-expanding chemical libraries. Moreover, while large-scale docking is advantageous against well-established protein binding sites, the effect of highly ranked artifacts is amplified by targeting unconventional protein sites, which are gaining interest in a broad range of diseases. Together, these limitations pose a significant challenge to realizing the therapeutic potential of available chemical space. My proposed program aims to build new tools for the efficient sampling of large chemical libraries, leveraging machine learning to accelerate and improve physics-based methods. Machine learning has exhibited enormous potential in accelerating the discovery process in many scientific areas, including chemistry, with demonstrated applications in quantum and classical mechanics calculations. Drawing inspiration from these recent advances, the main objectives of this proposal are: i)To develop machine learning-based accelerators for biomolecular simulations, with expanded applicability and more accurate prediction of ligand-receptor interactions; and ii)To build predictors of virtual screening performance, thereby guiding efficient targeting of unconventional and challenging binding sites. The program will be carried out by a diverse team of undergraduate and graduate HQP, providing them with training in central areas of modern computational chemistry, Artificial Intelligence (AI) and medicinal chemistry. The multidisciplinary nature of my program will provide the students with opportunities to fruitfully interact with academic and industrial groups active in a broad range of disciplines, from computer science to biology. This exposure will ensure that students develop an interdisciplinary research approach, preparing them for successful careers in chemistry software companies, the biotech and pharma industries, in health research, and in the growing AI sector. Our effort will democratize access to the chemical space to decipher its emerging yet elusive role in drug discovery. The program will ultimately augment the arsenal of tools available to the broad drug discovery community, benefitting the Canadian pharmaceutical and software sectors as well as accelerating the development of life-saving therapeutics for Canadians.