The reductionist paradigm insists that the physicochemical behavior of proteins can eventually be simulated if enough details of the molecular physics are included in the model. At the moment, however, completely atomistic simulations are very computationally time-consuming for even an individual biomolecule and are ill-suited for surveying the behavior of entire families of biomolecules. Fortunately, many (but not all!) of the aspects of biomolecular dynamics are sufficiently collective in their origin so that the relevant dynamics can be understood with coarse-grained models that capture the global characteristics of the energy landscape while still perhaps being inaccurate in some fine grained details of the landscape. Although we may sacrifice precision when using coarse grained models, they allow us to sample the entire configuration space, allowing accurate determinations of free energy surfaces and accurate statistics that relate to experimental observables, difficult to access with present day all-atom simulations.
Showing the most recent 10 out of 108 publications
