Peter V. Henstock
Pfizer
Over the past several years, computer science research has significantly advanced the field of computational biology. Computer science approaches can certainly impact a range of challenging problems in computational chemistry and drug discovery but have attracted far less attention. This talk will provide a general overview of the drug discovery process and outline a few potential research areas. The talk will focus on the problem of compound library design that aims to identify optimal compounds from a large set through several iterations of guided sampling. Techniques include forming compound representations, clustering, diversity sampling, and training predictive models.