G protein-coupled receptors (GPCRs) are the largest family of proteins in the human genome, and are the targets of a large fraction of modern pharmaceuticals. Although historically it has been difficult to obtain high resolution structural data on integral membrane proteins, in the last few years there has been an explosion of new information, with opsin, beta-2- and beta-1-adrenergic receptor, and A2A adenosine receptor joining rhodopsin as the only GPCRs with structures known at high resolution. My lab has investigated several GPCRs --- rhodopsin, beta-2-adrenergic receptor, CB2 cannabinoid receptor -- using microsecond-scale all-atom molecular dynamics simulations. These simulations remain among the longest ever performed, made possible by our longstanding collaboration with IBM Research.
Recently, we have complemented this work with investigations using elastic network models (ENMs). ENMs are models for representing protein fluctuations where the details of the all-atom forcefield are replaced with a set of harmonic springs connecting neighbouring residues; within this representation, the protein's motions can be solved in one step using normal mode analysis. The result is a computationally inexpensive yet surprisingly effective approach for understanding protein fluctuations.