Semiflexible Polymer Dynamics with a Bead-Spring Model

TL;DR

This paper develops an analytical and simulation-based study of semiflexible polymer dynamics, specifically double-stranded DNA, using a bead-spring model that aligns well with experimental and phenomenological data.

Contribution

It introduces a bead-spring model for semiflexible polymers, derives analytical expressions for key dynamical properties, and validates them with simulations, showing linearized dynamics are sufficient for short DNA segments.

Findings

Eigenmodes match established semiflexible polymer phenomenology.

Analytical correlation functions agree with simulation results.

Identifies a t^{7/8} nonlinear regime in end-to-end vector displacement.

Abstract

We study the dynamical properties of semiflexible polymers with a recently introduced bead-spring model. We focus on double-stranded DNA. The two parameters of the model, $T^*$ and $\nu$, are chosen to match its experimental force-extension curve. The bead-spring Hamiltonian is approximated in the first order by the Hessian that is quadratic in the bead positions. The eigenmodels of the Hessian provide the longitudinal (stretching) and transverse (bending) eigenmodes of the polymer, and the corresponding eigenvalues match well with the established phenomenology of semiflexible polymers. Using the longitudinal and transverse eigenmodes, we obtain analytical expressions of (i) the autocorrelation function of the end-to-end vector, (ii) the autocorrelation function of a bond (i.e., a spring, or a tangent) vector at the middle of the chain, and (iii) the mean-square displacement of a tagged bead in the middle of the chain, as sum over the contributions from the modes. We also perform simulations with the full dynamics of the model. The simulations yield numerical values of the correlation functions (i-iii) that agree very well with the analytical expressions for the linearized dynamics. We also study the mean-square displacement of the longitudinal component of the end-to-end vector that showcases strong nonlinear effects in the polymer dynamics, and we identify at least an effective $t^{7/8}$ power-law regime in its time-dependence. Nevertheless, in comparison to the full mean-square displacement of the end-to-end vector the nonlinear effects remain small at all times --- it is in this sense we state that our results demonstrate that the linearized dynamics suffices for dsDNA fragments that are shorter than or comparable to the persistence length. Our results are consistent with those of the wormlike chain (WLC) model, the commonly used descriptive tool of semiflexible polymers.