Protein Design with Deep Learning: Progress, Challenges, and Next Steps

The human proteome comprises tens of thousands of proteins, each tailored for a specific function by the selective pressures of evolution. The field of protein design seeks to develop proteins with new or enhanced functions at will, ultimately bypassing the evolutionary clock. In this talk, we discuss general machine-learning methods for accelerating the crucial components of protein design, with a particular focus on the underlying models.
We begin with an overview of the protein design field and discuss traditional approaches. We then focus on two important subtasks, fixed-backbone design and protein-protein docking, and critique deep learning approaches with an emphasis on the modeling considerations that attribute to their success. Finally, we highlight some limitations and outstanding challenges in the field and potential next steps.