Abstract
Drug development faces persistent challenges, with high attrition rates and unexpected adverse effects contributing to clinical trial failures. The recent convergence of large-scale biobanks, multi-omics data and computational methods, including machine learning, has led to advances in genetics-driven drug discovery, offering new opportunities to refine target selection and reduce late-stage risk. Integrating multiple lines of evidence centred on human genetics within a probabilistic framework enables the systematic prioritization of drug targets, prediction of adverse effects, and identification of drug repurposing opportunities. In this Review, we explore how these integrative approaches can address unmet clinical needs in diverse disease contexts, focusing on complex diseases.