The overall picture of molecular recognition covers the thermodynamic and kinetic properties of molecular systems. Regarding thermodynamics, binding affinity or free energy, which can be decomposed into enthalpy and entropy, determines the strength of binding. Binding kinetics, the association and dissociation rate constants, describe the rates for molecular binding and unbinding. In this thesis, I carried out various molecular mechanics modeling tools to understand binding thermodynamics and kinetics of host-guest systems and protein-ligand systems.
The work includes both novel method development and applications using existing tools. Here we computed the binding free energy of the surfactant-silver nano-plate complexes using the M2 method and to understand how a capping ligand helps grow shaped nanomaterials. Several factors were determined to be crucial, and our findings allow rational design of capping surfactants in addition to the magic compound citrate. In addition, a molecular dynamics (MD) and docking were applied to find the inhibitor binding site on protein TIM44 to assist cell biology studies. I examined the inhibitor selectivity of wildtype TIM44 and its mutants using energy calculations.
The binding/unbinding processes are carefully investigated for β-cyclodextrin and its guests to thoroughly study the thermodynamics and kinetics of binding. I applied MD simulations, post-analyzed the MD trajectories and developed and implemented methods such as the cell method to compute binding enthalpy, solute entropy and solvent entropy. Notably, the latter two entropy components are known to be very challenging to obtain computationally. I also computed the kinetic rate constants and investigated the influence of force fields used. It is the first work that illustrates all entropy and enthalpy components of binding in great details.
Realizing how crucial conformational sampling can be in answering fundamental scientific questions and valuable applications, a novel method to accelerate searching molecular conformations and ligand dissociation pathways was developed. The method uses a multi-layer internal coordinate and an internal PCA search algorithm to post-analyze a given MD trajectory and then guide motions of a molecular system. This algorithm is able to find possible dissociation pathways with high efficiency. Five examples are discussed to illustrate the functionality and capacity of this new method.