Analytical Framework to Uncover Structural Signatures from All-Atom Molecular Dynamics Simulations

Tang, Wai Shing

Full Metadata

Overview

Year:: 2022
Contributor:: Tang, Wai Shing (creator); Crawford, Lorin (Advisor); Rubenstein, Brenda (Reader); Tang, Jay (Reader); Brown University. Department of Physics (sponsor)
Genre:: theses
Subject:: Computational biology; Gaussian processes; Molecular dynamics; Machine learning--Statistical methods; computational physics; Molecular Dynamics Simulation
Extent:: 16, 158 p.

Files

Description

Abstract:: Structural features of proteins often serve as signatures of their biological function and molecular binding activity. Elucidating these structural and dynamical features is essential for understanding underlying biophysical mechanisms. Significant efforts in the past decade have resulted in highly accurate all-atom protein force fields for molecular dynamics (MD) simulations of proteins. These simulations provide new insights into the underlying physics of both folded proteins, which are easy to characterize by one or a few native states, and intrinsically disordered proteins (IDPs), which are structurally heterogeneous and challenging to characterize by experiments alone. However, state-of-the-art protein force fields still display significant discrepancies in MD simulation compared to experiments. Moreover, given accurate structural ensembles obtained from MD simulation, existing methods do not have the statistical capability to robustly identify geometric and dynamic variations in these simulated protein ensembles or describe the molecular interactions underlying the biophysical behavior of the protein systems. In this thesis, I present the analytical approaches for optimizing the accuracy of MD simulations and extracting structural signatures from the simulated ensembles. In the following chapters, I present (1) a refinement to the all-atom molecular dynamic simulation force field to improve the representation of helicity for intrinsically disordered proteins, (2) a statistical approach to quantify and visualize the molecular interactions underlying the liquid-liquid phase separation of intrinsically disordered protein systems, and (3) a novel topological data analytic approach for discovering biophysical signatures in protein dynamics.
Notes:: Thesis (Ph. D.)--Brown University, 2022

Content

Citation

Tang, Wai Shing, "Analytical Framework to Uncover Structural Signatures from All-Atom Molecular Dynamics Simulations" (2022). Physics Theses and Dissertations. Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:ybkvqm5d/

Relations

Collection:

Physics Theses and Dissertations

Theses and Dissertations for the Physics department.
...