Using AlphaFold protein models for structure-based virtual screening of chemical libraries
Structural Biology seminar by Albert Ros-Lucas, Barcelona Institute for Global Health
Hosted by EMBL-EBI
Abstract
This webinar uses research on Chagas disease to explain how predicted protein structures can support drug discovery when experimental structures are unavailable. In the example study, AlphaFold models of Trypanosoma cruzi proteins supported a virtual screen of roughly 30,000 compounds; subsequent experiments identified antiparasitic activity in two approved drugs. The session introduces AlphaFold confidence measures and ways to obtain models, then follows a screening workflow through binding-pocket prediction, chemical-library selection, docking, and interpretation of results. It discusses difficulties encountered when docking against predicted structures and considerations for adapting the workflow to other targets. The material is intended for doctoral students, early-career researchers, and other drug-discovery professionals.
Topics
Related seminars
AI-based structural modelling of host-pathogen protein interactions
More on alphafold
Automated X-ray fragment and ligand screening for therapeutic innovation in infection biology
Related research
From transcriptomics to structural modelling with AlphaFold: analysis of the central response to stress in bacteria using machine learning and structural bioinformatics
More on alphafold