Graduation Year

2023

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Chemistry

Major Professor

Arjan van der Vaart, Ph.D.

Committee Member

Brian Space, Ph.D.

Committee Member

Ioannis Gelis, Ph.D.

Committee Member

Sagar Pandit, Ph.D.

Keywords

Computational Chemistry, Enhanced Sampling, Molecular Dynamics, Simulation

Abstract

This work uses molecular dynamics simulations and enhanced sampling methods to investigate both the qualitative and quantitative properties of biomolecular systems. Conformational dynamics are critical in the study of biomolecular systems. Application of computer simulations to study dynamics of biomolecular systems allows for motions to be readily investigated at a resolution not possible in experimental techniques. In this work, the dynamics of uracil damaged and undamaged DNA is studied via molecular dynamics simulations. Uracil results in DNA from either misincorporation in the replication process or through spontaneous deamination of cytosine, and its presence in DNA can lead to mutations in the genome and even influence molecular evolution. The role of nearest neighbors and their influence on the dynamics around uracil lesions are of interest to better understand the role of deformability associated with lesions in different sequence contexts. Experimental measurements of flexibility and kinetics with simulated flexibility measurements are compared, and a further analysis regarding the link between behavior of uracil- damaged DNA and its undamaged counterpart are investigated. In addition to studying the conformational dynamics of DNA, this study also delves into the difficult and computationally demanding process of calculating conformational free energy differences by introducing the focused confinement method. Accurately and efficiently sampling a given potential energy landscape becomes progressively complex as biomolecular systems increase in size, with difficulties such as large energy barriers, intermediate states, and increased time scales for slow modes of motion impeding sampling. While extant enhanced sampling may calculate conformational free energies with some accuracy, many are unfortunately limited due to thevi aforementioned obstacles; such methods, and their limits, are explored in the text. This work introduces the focused confinement method which helps solve many of the problems associated with calculating conformational free energy differences. The focused confinement method builds off traditional confinement by harmonically restraining only the atoms that actively undergo a conformational change. Restraining only the conformationally active atoms requires fewer restraining simulations, resulting in reduced computational costs. Results show that the focused confinement is independent of the complexity of the conformational change and is useful for large explicitly solvated systems. Accurate calculations for desolvation free energies are possible with free energy perturbation due to sufficient rigidification of carefully defined restraints on conformationally active atoms. This work introduces a general partitioning procedure to help optimally define conformationally active atoms so the free energy of desolvation and the free energy between the mixed harmonic-anharmonic states are also optimized. Results show that free energies are independent of partitioning, thus conformationally active atoms can be chosen in a minimal set to focus confinement efficiently. Derivation and results of focused confinement will be explored in text.

Included in

Chemistry Commons

Share

COinS