Conformation sampling enumeration

First, the current state of affairs is remarkably similar to that of the field of computational molecular dynamics 40 years ago. While the basic equations are known in principle (as we shall see), the large number of unknown parameters makes realistic simulations essentially impossible. The parameters in molecular dynamics represent the force field to which Newton s equation is applied the parameters in the CME are the rate constants. (Accepted sets of parameters for molecular dynamics are based on many years of continuous development and checking predictions with experimental measurements.) In current applications molecular dynamics is used to identify functional conformational states of macromolecules, i.e., free energy minima, from the entire ensemble of possible molecular structures. Similarly, one of the important goals of analyzing the CME is to identify functional states of areaction network from the entire ensemble of potential concentration states. These functional states are associated with the maxima in the steady state probability distribution function p(n i, no, , hn). In both the cases of molecular dynamics and the CME applied to non-trivial systems it is rarely feasible to enumerate all possible states to choose the most probable. Instead, simulations are used to intelligently and realistically sample the state space. 

This procedure, however, is impractical because the generation of the representative set of single-chain conformations in the external field via importance sampling is computationally more expensive than partial enumeration and one would have to perform a computer simulation in each iteration cycle of the self-consistent adjustment of fields and densities. 

---