Graduation Year

2024

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Chemistry

Major Professor

H. Lee Woodcock, Ph.D.

Committee Member

James Leahy, Ph.D.

Committee Member

Ioannis Gelis, Ph.D.

Committee Member

Yu Chen, Ph.D.

Keywords

computational chemistry, covalent docking, de novo drug design, qm/mm, molecular docking

Abstract

Modern drug discovery, the process of identifying or developing novel therapeutic compounds to treat a disease, increasingly relies on computational methods to make the process faster and more efficient. One of the most useful tools in the computer-aided drug design toolbox is molecular docking, the simulation of a protein-ligand complex to determine ligand binding affinity. The results from molecular docking help inform the overall drug discovery process on which compounds have more potential to be potent inhibitors, thus saving time and resources during the later and more expensive stages of drug discovery. Two important considerations of modern docking methods are flexibility and timing. Adequate sampling should be conducted on the protein-ligand complex to thoroughly incorporate induced fit effects and ensure that relevant binding modes can be obtained. At the same time, conformational exploration must be limited to keep run time and computational cost low when in the context of virtual screening. To this end, we have developed a novel flexible docking method, dubbed CIFDock, which uses enhanced dynamics, active site mutations, and rotamer libraries to capture protein-ligand flexibility and induced fit effects. Modifications to the protocol also enable covalent inhibitor screening and de novo ligand design, making CIFDock a powerful tool for drug discovery. The introduction will briefly cover the concept of drug discovery, the computational tools used in computer-aided drug design, and an overview of molecular docking. Chapter 2 will describe the CHARMM-based flexible docking method (CIFDock) developed in our lab and highlight its benefits over modern commercial alternatives. Chapter 3 will explain the first major modification to CIFDock - the use of hybrid QM/MM minimizations, which enable covalent inhibitors to be screened with our docking method. Chapter 4 will introduce another major modification to CIFDock - the incorporation of various cheminformatics programs into a de novo design subroutine, which enables fragment-by-fragment ligand design that is custom fit to a target binding site through simultaneous docking. By the end of this dissertation, I hope to enlighten readers on our novel docking method and its powerful subsequent modifications which make it a robust tool for drug design that is freely available to any interested researcher.

Download

Share

COinS