Quantitative models of biomolecular hydration thermodynamics
MetadataShow full item record
This thesis explores the use of cell theory calculations to characterise hydration thermodynamics in small molecules (cations, ions, hydrophobic molecules), proteins and protein-ligand complexes. Cell theory uses the average energies, forces and torques of a water molecule measured in its molecular frame of reference to parameterise a harmonic potential. From this harmonic potential analytical expressions for entropies and enthalpies are derived. In order to spatially resolve these thermodynamic quantities grid points are used to store the forces, torques, and energies of nearby waters which giving rise to the new grid cell theory (GCT) model. GCT allows one to monitor hydration thermodynamics at heterogeneous environments such as that of a protein surface. Through an understanding of the hydration thermodynamics around the protein and particularly around binding sites, robust protein-ligand scoring functions are created to estimate and rank protein-ligand binding affinities. GCT was then able to retrospectively rationalise the structure activity relationships made during lead optimisation of various ligand-protein systems including Hsp90, FXa, scytalone dehydratase among others. As well as this it was also used to analyse water behaviour in various protein environments with a dataset of 17 proteins. The grid cell theory implementation provides a theoretical framework which can aid the iterative design of ligands during the drug discovery and lead optimisation processes, and can provide insight into the effect of protein environment to hydration thermodynamics in general.