UC Irvine

The development of a viable therapeutic and getting a drug approved to help patients is an expensive process in terms of both resources and time, even in modern drug discovery. It can take about 7 to 10 years for a therapeutic to be approved and brought to market for prescription, and the entire process can cost up to several billions of dollars. Computational methods can be used to assist with HTS (high-throughput screening), hit identification, and optimization -- and are poised to make the process more efficient by reducing time and cost of drug development. Particularly, binding free energy (BFE) calculations based on molecular dynamics simulations has gained traction as a powerful tool to provide insight to the dynamics of a target and ligand, the ligand binding mode(s), prioritizing molecules for synthesis via binding affinity, and helping in accelerating small-molecule drug discovery. There are reported success in academia and industry as well as some known limitations and reported problems, requiring further investigation and a need for improvement and innovation.

In this dissertation, I present my work both on a series of blind challenges, and structure-based inhibitor design. Particularly, I present my work on the Statistical Assessment of the Modeling of Proteins (SAMPL) -- a series of blind challenges -- where I benchmark the accuracy of binding free energy predictions of small molecules to macrocyclic supramolecules, construct workflows and add automation using existing toolkits to setup systems for binding free energy calculations and apply them to SAMPL host-guest systems. In addition, I apply computational tools to help guide structure-based design of inhibitors. Overall, my work comprises of applying computational tools to study biologically relevant systems and development of tools to aid in early stage therapeutic development, aimed to help improve the quality of life.