Towards optimal chemical space search with generative virtual screening

Event details

Date	03.03.2026
Hour	15:15 › 16:15
Speaker	Morgan Thomas is a Postdoctoral Fellow at Khalifa University in Abu Dhabi, where he leads machine learning in medicine initiatives with Prof. Andreas Bender. Following his experience on the AstraZeneca Graduate Programme, he received his PhD from the University of Cambridge sponsored by Nxera Pharmaceuticals. He conducted his first postdoctoral fellowship with Prof. Gianni de Fabritiis at Universitat Pompeu Fabra supported by a Johnson & Johnson Innovative Medicine Grant. His research focuses on machine learning methods to automate and improve drug discovery, particularly through generative AI, structure-based principles, and benchmarking frameworks. He consistently works at the interface of academia and industry, grounding his research in practical application.
Location	Online
Category	Conferences - Seminars
Event Language	English

Efficiently searching virtual chemical space is central to discovering new chemical matter. Traditionally, virtual screening of enumerated libraries and related approximations has been used to navigate accessible subsets of chemical space. More recently, generative AI has emerged as a strategy for implicitly exploring vastly larger chemical spaces. In this talk, I compare these paradigms for chemical space search, focusing on computational efficiency, practical progressability, and evaluability. I will share recent work on improving the sample efficiency of reinforcement learning–based generative methods and introduce new benchmarks that stress-test current approaches under extreme search requirements, clarifying how close we are to truly optimal chemical space exploration.