INTRODUCTION

The routine use of quantum mechanics (QM) in all phases of in silico drug design is the logical next step in the evolution of this field. The first principles nature of QM allows it to systematically improve the accuracy of the description of the nature of the interactions between molecules. Moreover, the systematic way in which one can approach the use of QM methods to solve chemical and biological problems is quite appealing, but the practical use of many of the appealing features of QM in in silico drug design applications is still to be realized in large part because of computational limitations. In recent years it has become clear that classical potential functions are being pushed to their limits and as many pitfalls of using them are coming to light, one is tempted to explore the use of QM procedures. This is a somewhat naïve view, however, because one of the main observations of a large body of computational work has shown that sampling of relevant conformational states can be as important as providing an accurate representation of an inter-or intramolecular interaction. Hence, even as QM becomes a routine tool used to calculate the energy of individual states of a biological system, one still faces the daunting task of sampling relevant conformational space, which, in our view, will for the near term be largely confined to classical models.

Our goal in producing this book is to provide a broad overview of the most important approaches used in protein- and ligand-structure-based drug design. Beyond this we aim to illustrate how these approaches are currently being applied in drug discovery efforts. We hope this book will be a useful resource to practitioners in the field, as well as a good introduction for researchers or students who are new to the field. We believe it provides a snapshot of the most important trends and capabilities in the application of modeling and structural data in drug discovery.

Since the 1990s the role of structure and modeling in drug discovery has grown enormously. There have been remarkable scientific advances in both the experimental and computational fields that are the underpinnings of modern drug design. For example, x-ray capabilities have improved to the point that protein structures are now routinely available for a wide range of protein targets. One only need look at the exponential growth of the Protein Databank (RCSB) for evidence. Tremendous strides have been made in all aspects of protein structure determination, including crystallization, data acquisition, and structure refinement. Modeling has made similar gains. Recent years have brought more realistic force fields, new and more robust free-energy methods, computational models for absorption/distribution/metabolism/excretion (ADME)-toxicity, faster and better docking algorithms, automated 3D pharmacophore detection and searching, and very-large-scale quantum calculations.