Artificial Intelligence to Search for Drug Leads in Gigascale Chemical Spaces as a way to Fight Antimicrobial Resistance in Malaria

2026-07-22

Malaria, a global infectious disease, is facing a critical issue of drug efficacy loss due to the alarming rise of antimicrobial resistance (AMR). To combat this, a promising strategy is the discovery of cell-active drug leads hitting new targets, such as P. falciparum dihydroorotate dehydrogenase (PfDHODH). However, two main roadblocks prevent us from achieving this important goal: 1) many molecules with potent activity on a protein target are later found to be inactive in pathogen cultures for various reasons, 2) identifying novel drug leads with potent activity on both protein and phenotypic targets has traditionally been a lengthy and costly process. A timely opportunity leveraging the fast-growing volumes of bioactivity and protein structure data to guide virtual screening. Indeed, Artificial intelligence (AI)-guided virtual screening has boosted hit rates and potencies in both molecular and phenotypic activity assays on other human pathogens. This multidisciplinary project will therefore investigate cutting-edge AI models to guide structure-based virtual screening against the molecular target PfDHODH. This project will use the best of these AI models to screen drug-like ultralarge libraries for novel molecules with potent activity on the considered target in a fast and cost-effective manner. To predict which of these target-active molecules are also cell-active, we will investigate and use AI models to guide phenotypic virtual screening against reference (Pf3D7) and drug-resistant (PfDd2) pathogen cultures.

United Kingdom (UK)

2026-01-01