Title: Expanding the druggable proteome with cryptic pockets and allostery
Bio:
Dr. Greg Bowman is the Louis Heyman University Professor at the University of Pennsylvania. He and his lab devise new ways to design therapeutics and interpret genetic variation in areas ranging from viral infections like Ebola to Alzheimer’s and cardiac disease. A major emphasis is on identifying cryptic pockets and allostery by mapping out the ensemble of different structures that a protein adopts. To build and exploit these maps, the group combines biophysical experiments, physics-based simulations, and machine learning. Greg is also the Director of the Folding@home distributed computing project, which harnesses the computational power of citizen scientists from across the globe to enable computer simulations on an unprecedented scale.
Abstract:
Many proteins are thought to be difficult drug targets, or even outright “undruggable,” because their structures lack pockets that appear amenable to drug discovery or their functional sites are conserved, making specificity difficult to achieve. Considering protein dynamics could alleviate these concerns in many cases, greatly expanding the druggable proteome. For example, ‘cryptic’ pockets that are absent in known structures of proteins but form due to protein dynamics could provide a means to target proteins thought to lack druggable pockets. Furthermore, allosteric sites, whether cryptic or not, could provide a means to specifically target subsets of proteins with shared functional sites. However, it has been difficult to assess or exploit these dynamical features due to the inherent difficulty in identifying them. Here, I will discuss progress from my lab on leveraging machine learning to identify and target cryptic pockets and allostery.
Keywords: protein dynamics, drug discovery
Host: Prof. Xuhui Huang