AI-assisted fragment-based drug discovery of SARS-CoV-2 macrodomain binders validated by NMR and X-ray crystallography

Fragment-based drug discovery (FBDD) has become an effective approach for exploring chemical space by employing small, low-affinity binders to facilitate development of lead compounds. Strategies used for transforming these initial weak binders into more potent inhibitors have included fragment merging and linking to enable their transformation into high-affinity compounds. Recently, the integration of artificial intelligence (AI) and machine learning (ML) has further expedited this process by supporting structure-based optimization and generative compound design. Here, we demonstrate this AI-assisted FBDD approach as a proof of concept by applying it to the SARS-CoV-2 Macrodomain (Mac1), a conserved viral protein involved in immune evasion and ADP-ribose metabolism. Leveraging extensive structural data and previously identified fragments, we employed deep learning and docking to design novel Mac1 binders. Several compounds were synthesized and validated via NMR and X-ray crystallography, demonstrating improved binding affinities. This investigation highlights the synergistic advantages of combining AI with FBDD to streamline the design and prioritization of candidate molecules, providing a data-driven framework for discovering new Mac1 inhibitors and guiding future antiviral drug development.