Dynamic Allosteric Biomolecular Design Using Artificial Intelligence
Kelly M. Thayer, David L. Beveridge, David R. Langley · The FASEB Journal · 2019
Protein‐ligand binding mediated by an allosteric effector can be considered from the vantage of dynamic macromolecular structure. Recently, we have developed a methodology, MD‐based Markov State Models (MD‐MSMs) making use of this view. We envision gaining insight into the workings of allosteric systems, and applying those principles to allosterically engineer biomolecules as a new class of therapeutics. Such allosteric regulators hold promise for modulating currently “undruggable” targets. In our approach, Molecular dynamics simulations (MD) generate the Boltzmann ensemble of conformations of a system. Using machine learning to process trajectories, the number of conformational substates can be learned from the conformations of the system (Figure 1A). Network theory captures the dynamic interchange between conformations (Figure 1B). Our recent findings from the application of MD‐MSMs to allosteric proteins (Figure 1C) and principles for engineering allosteric effectors will be presented. Support or Funding Information NIH R15 GM128102 to KMT This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .