I'm not an expert in molecular dynamics simulations. All of my knowledge comes from hearing my girlfriend, a graduate student in Dr. Carol Post's lab at Purdue, give talks on her research using and developing this method. I do have an appreciation for what can and cannot be accomplished by these techniques. I am blown away by the results presented by the last two National Lecturers, Klaus Schulten and David Shaw.
My own research utilizes electron paramagnetic resonance spectroscopy to study membrane protein dynamics and conformational changes. Using this method, I use available crystal structures and homology models to guide the placement of spin probes via site-directed mutagenesis. These probes then allow me to study dynamics and conformational changes in solution and in the context of protein function, providing me low-resolution information to supplement high-resolution structures gained from crystallography and EM.
The key here is low-resolution: I can study changes in and across specific regions and interpret these in the context of structural rearrangements and quaternary assemblies. Comprehensive information about the movement of every atom would provide a more complete picture of protein function. Of course, several different biophysical methods - EPR, NMR, IR, CD, mass spec - exist and can in tandem provide a lot of this information. These come with their own challenges to perform and interpret and aren't exactly high throughput.
What I take away from these long, beautiful simulations performed by David Shaw's group is that they provide testable hypotheses for protein functional studies and guide drug design in a dynamic environment. I was reminded during his talk of the simulations of the bacterial voltage-gated potassium channel KcsA, several hundred microseconds long, that showed asymmetric translations of the voltage-sensing domains leading to channel closure via a series of partially open and closed states. Earlier in the day at the poster session, I spoke to Dorothy Kim, a postdoctoral researcher at Weill Cornell Medical College, who studies KcsA dynamics using a solution NMR technique, heteronuclear single quantum coherence (HSQC). By reconstituting KcsA in natural membrane environments and labeling key interfacial histidines, Kim demonstrated KcsA samples an ensemble of states going from the inactive to the active channel. These data provide experimental evidence that dynamics occur even in closed conformations, which cannot be deduced from static crystal structures.
-Satchal Erramilli