Artificial intelligence for RNA–ligand interaction prediction: advances and prospects

Accurate prediction of RNA–ligand interactions is vital for understanding biological processes and advancing RNA-targeted drug discovery. Given their complexity, artificial intelligence (AI) is revolutionizing the study of RNA–ligand interactions, offering insights into the complex dynamics and therapeutic potential of RNA. In this review, we highlight advances in AI-driven RNA–ligand binding site identification, structure modeling, binding mode and binding affinity prediction, and virtual screening (VS). We also discuss key challenges, such as data set scarcity and modeling RNA flexibility. Future directions emphasize integrating cutting-edge AI techniques with physics-based models and expanding experimental data sets to enhance RNA–ligand interaction predictions.