DiffDock-Glide: A Hybrid Physics-Based and Data-Driven Approach to Molecular Docking

Recent years have seen a rise in applications of deep learning to problems in the molecular sciences. Among them, the diffusion model DiffDock stands out as a method for docking small molecules into protein binding sites. But DiffDock struggles to compete with conventional docking methods, especially for targets outside its training set. We develop a hybrid model called DiffDock-Glide which addresses some shortcomings of deep learning docking methods: it uses a modified generative process to generate samples within a binding pocket, and the confidence model is replaced with Glide's postdocking minimization pipeline. We evaluate DiffDock-Glide on the PoseBusters data set and show improved sampling of near-native poses, especially for sequences without homologues in the training set. We also evaluate DiffDock-Glide's performance in virtual screening of compounds from the DUD-E data set against receptor structures generated by AlphaFold2 and report enrichment values that broadly surpass those from traditional Glide.