Advancing Quantum Mechanical Methods for Non-Covalent Interaction Analysis Across System Scales

Abstract

The discovery and design of new drugs remain critical challenges in modern chemistry and biology. Traditional drug development is notoriously expensive and time-consuming, often requiring over a decade and billions of dollars to bring a single drug to market, with a high failure rate due to inefficacy or safety concerns. Computational chemistry has emerged as a pivotal tool to enhance efficiency and success rates in drug discovery by simulating molecular interactions and predicting compound properties in silico. This thesis aims to rigorously investigate reactivity and non-covalent interactions in systems of varying sizes by developing and applying advanced computational methods. Specifically, we focus on the Interacting Quantum Atoms (IQA) energy decomposition method and we pair it with the Relative Energy Gradient (REG) approach. This approach allows for the extraction of chemical insights by identifying system fragments that mirror the behaviour of the entire system. We extend REG to include multipolar electrostatics (REG-MULTI), enhancing its applicability to electrostatic environments prevalent in protein structures. We apply the REG-IQA method to molecular balances-experimentally designed organic molecules that probe specific non-covalent interactions. We also utilise REG-MULTI and REG-IQA to investigate XH-pi interactions through prototype small systems representative of common interactions in protein environments. Given IQA's large computational expense, we develop a rigorous approach combining quantum chemistry with computational geometry to reduce computational costs for large DFT clusters. This approach is then applied to study G-protein-coupled receptors in collaboration with experimentalists. REG-IQA uncovers mechanistic insights into these essential enzymes responsible for cellular signalling and function. Overall, this work demonstrates that advanced computational techniques like REG-IQA can significantly enhance both qualitative and quantitative interpretations of experimental data. By providing deeper insights into molecular interactions and reactivity, the methodologies developed herein offer a robust framework for future applications in drug discovery and the study of complex biological systems. This thesis not only advances computational methods but also strengthens the synergy between theoretical predictions and experimental validations in chemistry.

Date of Award	16 Dec 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Jonathan Waltho (Main Supervisor) & Paul Popelier (Co Supervisor)