The University of Southampton
University of Southampton Institutional Repository
Warning ePrints Soton is experiencing an issue with some file downloads not being available. We are working hard to fix this. Please bear with us.

Enhancing conformational sampling by modifying the underlying velocity distribution: Digitally filtered Hybrid Monte Carlo

Enhancing conformational sampling by modifying the underlying velocity distribution: Digitally filtered Hybrid Monte Carlo
Enhancing conformational sampling by modifying the underlying velocity distribution: Digitally filtered Hybrid Monte Carlo
The rugged energy landscape of proteins often leads conventional molecular dynamics simulations to get stuck in local minima leading to inefficient sampling and slow convergence. Two main methods exist to improve sampling. First, by smoothing the underlying energy surface thereby encouraging barrier crossing, and second, by modifying the simulation velocities. There are many examples in the literature of the former, but velocity modification is a comparatively less well exploited approach. In a method previously developed in our group called Reversible Digitally Filtered Molecular Dynamics (RDFMD), we have been able to induce conformational changes in proteins by amplifying low-frequency molecular vibrations via application of a digital filter to the velocity set [1]. However, the application of a digital filter to the molecules, disrupts the equilibrium of the system, leading to incorrect ensemble averages. To sample a statistical ensemble and still benefit from the application of the digital filter a novel method called Digitally Filtered Hybrid Monte Carlo (DFHMC) is proposed. This method builds on the work of Momentum Enhanced Hybrid Monte Carlo (MEHMC) [2]. Low frequency motions are selectively enhanced via application of a specifically designed Digital Filter (as in RDFMD), but equilibrium is maintained using a Hybrid Monte Carlo approach. In this thesis, the theory behind the DFHMC and the integration of the method into several optimised molecular dynamics packages is explained. The capability of sampling from the canonical ensemble and the enhancement in energy barrier crossing, and hence convergence, is demonstrated on a system with double well potential and with a curved transition path. As a model for the dihedral motions in proteins, the application of the DFHMC method to a single alanine dipeptide molecule in explicit and implicit water is discussed. We observed a reduction in sampling efficiency in DFHMC when the molecule has centre of mass rotation. Ideas on how to potentially overcome the rotation problem is discussed. And the Riemannian Manifold Hamiltonian Monte Carlo method [3] is introduced to investigate the application of the method to molecular systems.
University of Southampton
Pervane, Can Simon
a691a5d7-50b8-4dbf-a4d3-106cb6a39796
Pervane, Can Simon
a691a5d7-50b8-4dbf-a4d3-106cb6a39796
Essex, Jonathan
1f409cfe-6ba4-42e2-a0ab-a931826314b5

Pervane, Can Simon (2019) Enhancing conformational sampling by modifying the underlying velocity distribution: Digitally filtered Hybrid Monte Carlo. Doctoral Thesis, 166pp.

Record type: Thesis (Doctoral)

Abstract

The rugged energy landscape of proteins often leads conventional molecular dynamics simulations to get stuck in local minima leading to inefficient sampling and slow convergence. Two main methods exist to improve sampling. First, by smoothing the underlying energy surface thereby encouraging barrier crossing, and second, by modifying the simulation velocities. There are many examples in the literature of the former, but velocity modification is a comparatively less well exploited approach. In a method previously developed in our group called Reversible Digitally Filtered Molecular Dynamics (RDFMD), we have been able to induce conformational changes in proteins by amplifying low-frequency molecular vibrations via application of a digital filter to the velocity set [1]. However, the application of a digital filter to the molecules, disrupts the equilibrium of the system, leading to incorrect ensemble averages. To sample a statistical ensemble and still benefit from the application of the digital filter a novel method called Digitally Filtered Hybrid Monte Carlo (DFHMC) is proposed. This method builds on the work of Momentum Enhanced Hybrid Monte Carlo (MEHMC) [2]. Low frequency motions are selectively enhanced via application of a specifically designed Digital Filter (as in RDFMD), but equilibrium is maintained using a Hybrid Monte Carlo approach. In this thesis, the theory behind the DFHMC and the integration of the method into several optimised molecular dynamics packages is explained. The capability of sampling from the canonical ensemble and the enhancement in energy barrier crossing, and hence convergence, is demonstrated on a system with double well potential and with a curved transition path. As a model for the dihedral motions in proteins, the application of the DFHMC method to a single alanine dipeptide molecule in explicit and implicit water is discussed. We observed a reduction in sampling efficiency in DFHMC when the molecule has centre of mass rotation. Ideas on how to potentially overcome the rotation problem is discussed. And the Riemannian Manifold Hamiltonian Monte Carlo method [3] is introduced to investigate the application of the method to molecular systems.

Text
final thesis for award
Restricted to Repository staff only until 24 June 2022.
Available under License University of Southampton Thesis Licence.
Text
PTD with approval
Restricted to Repository staff only

More information

Published date: February 2019

Identifiers

Local EPrints ID: 447860
URI: http://eprints.soton.ac.uk/id/eprint/447860
PURE UUID: 099b0215-0a87-4a26-840d-8a42cc9fd4e7
ORCID for Jonathan Essex: ORCID iD orcid.org/0000-0003-2639-2746

Catalogue record

Date deposited: 24 Mar 2021 18:31
Last modified: 25 Mar 2021 02:34

Export record

Contributors

Author: Can Simon Pervane
Thesis advisor: Jonathan Essex ORCID iD

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×