Linna An, postdoc at University of Washington, describes the process of designing small molecules that bind to proteins. The canonical method involved generating scaffolds, designing the ligand docking, designing docks, designing filters, then finally testing in the lab. However, there are difficulties with finding ideal ligand designs because small molecules have such low tolerances. This problem can be solved using deep learning computation to improve every step of the design process. Linna’s group is incorporating AlphaFold 2, RosettaFold, and Rosetta to dramatically speed up small molecule discovery.