| Therapeutics, whether antibodies, proteins, small molecules, or nucleic acids, rely on recognizing their targets with high affinity and selectivity in order to be effective. This ability to discriminate one species from another is central in drug design and its various mechanisms have yet to be fully understood. Researchers are now looking into computational methods to gain an increased understanding of the processes involved in molecular recognition.; While there exists a vast array of approaches, this work focuses primarily on a rigorous, generalizable method in structure-based drug design, namely molecular dynamics and MMPBSA. This method enables detailed descriptions of receptor-ligand complexes at the atomic level, providing a quantitative insight into the molecular mechanisms of interaction.; Chapter 2 begins our investigation into molecular recognition and describes work predicting charge states of essential residues in a nucleic acid system. This study successfully predicted pKa shifts of residues in the intricate 50S ribosome active site and provided support for a proposed mechanism of peptidyl transfer. Chapter 3 furthers our comprehension of molecular association through energetic studies on a carbohydrate-protein complex. Successful estimates of the free energies of binding and successive decomposition of the binding determinants offered a detailed analysis of recognition and further enhanced our understanding of the importance of water molecules in association events.; Chapter 4 describes work investigating drug design from an alternative point of view. Instead of considering the receptor topology directly, we examined the databases that are commonly screened. In particular, we explored the development of a method that characterizes a small molecule database.; In addition to steric and chemical complementarity, protein motion also guides recognition. In our fifth chapter, we investigated the commonly used computational method to simulate motion, molecular dynamics. Using a high-resolution crystal structure, we followed atomic forces from an MD simulation which yielded characteristic frequencies for our protein in various simulation environments. To further analyze these methods, we also quantitated the atomic forces on particular atoms. We saw that instantaneous forces were chaotic and that, in general, the complexity of the simulation environment has a negligible effect on molecular motion. |
