Comparison of DNA force fields at a multiscale level : towards bottom-up modelling of chromatin

Computer simulations were performed to understand the dynamics of DNA molecules in presence of mono-valent cations. These simulations were run using different allatom DNA force fields to study their behaviour at a multiscale level, from the base-pair to the mesoscopic level. Serious problems were...

Full description

Saved in:
Bibliographic Details
Main Author: Minhas, Vishal
Other Authors: Lars Nordenskiöld
Format: Thesis-Doctor of Philosophy
Language:English
Published: Nanyang Technological University 2020
Subjects:
Online Access:https://hdl.handle.net/10356/142499
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Computer simulations were performed to understand the dynamics of DNA molecules in presence of mono-valent cations. These simulations were run using different allatom DNA force fields to study their behaviour at a multiscale level, from the base-pair to the mesoscopic level. Serious problems were found with the latest version of the CHARMM family of force fields, namely CHARMM36, as it distorted DNA structure and backbone at the microsecond timescale. Also, to achieve the long-term goal of modelling the folding of the chromatin fiber, we have built a multiscale model using systematic bottom-up approach based on structure based coarse-graining. At the first stage of bottom-up coarse-graining, all-atom and mesoatom representations of all principal chromatin components were bridged. To do this, the all-atom molecular dynamics trajectories obtained were bead-mapped to a coarse-grained representation and used for the calculation of the radial distribution functions, which were then used to derive the effective potentials using the Inverse Monte-Carlo method. These potentials were then validated in a coarsegrained molecular dynamics simulation of the original system. Based on these potentials, an excellent prediction of the dependence of DNA persistence length on salt concentration, in agreement with experiments was obtained, which validates the underlying CG-model. Once all the potentials are obtained for all the interactions, a large-scale CG simulation will be run using these potentials to model the NCP.