Advances and Continuing Challenges in Achieving Realistic and Predictive Simulations of the Properties of Organic and Biological Molecules

Abstract

Can the theoretical/computational chemist correctly predict the structures and free energies of molecular systems? It is an assumption of our research that classical molecular mechanics/dynamics is or can be made sufficiently accurate to accurately represent not only conformational (intramolecular) energetics but also nonbonded (both intramolecular and intermolecular) energetics. Even though these classical mechanical energy functions are not perfect representations of the BornOppenheimer surface, progress has been made during the last decade in improving them. If the force fields are proven accurate on rigid or degenerate systems such as liquids, where we can sample enough of the system to be confident of convergence, then they can be usefully applied to systems of greater complexity, where the main difficulty is sampling enough configurations of the system.1-3 Given the importance of an accurate molecular mechanical model, we first present the recent efforts to develop an improved one, dividing our discussion into (a) philosophy and approach to a molecular mechanical model, (b) testing the molecular mechanical model on simple systems, (c) nonadditive molecular mechanical models, and (d) the importance of long-range electrostatic effects in complex systems. We then turn to applications of such models to molecules of organic and biochemical interest. Here we highlight a number of themes: (a) solvation in simple and complex systems, (b) molecular recognition, (c) more complex systems, octanol and DNA, and (d) chemical reactions in solutions and in enzymes. We then conclude with perspectives for the future. [Accounts of Chemical Research articles are intended to emphasize research in the author’s lab; thus, we have been quite selective in our discussions. A recent exhaustive review of free energy calculations has attempted to note all the contributions to 1993.4 However, we note that the many exciting studies being carried out in other labs attest to the vitality of computational chemistry and biochemistry.]