(505e) Some Surprises Arising From Atomistic Simulation of Protein Dynamics On the Millisecond Timescale: Protein Folding and Conformational Change of Kinases and GPCRs

A key challenge in molecular simulation is reaching experimentally relevant timescales with models that are sufficiently detailed to quantitatively predict experiment.  Indeed, most atomistic simulations typically only reach the nanosecond to microsecond timescale, whereas experimentally relevant timescales for many phenomena are on the millisecond to second timescale.

We describe a new approach for simulating long timescale dynamics using Markov State Models (MSMs).  MSMs are a systematic sampling scheme to both get to long timescales as well as to gain insight from the results.  Indeed, this approach allows a cluster of GPUs to simulate a millisecond of aggregate dynamics and to use that data to quantitatively predict experiment, indicate the degree of robustness of results, and yield novel insights.  

Finally, I will demonstrate this method with applications to all-atom molecular simulations  on the millisecond timescale and beyond, with applications to protein folding, protein misfolding in Alzheimer's Disease, and protein conformational change in disease-relevant drug targets of GPCRs and kinases.