Memory-dependent quadratic action for fractal Brownian motion

Kirill Polovnikov, Sergei Nechaev, M. TammarXiv 2017 · arXiv

Abstract: In the framework of the $\beta$-model proposed by A. Amitai and D. Holcman, we analyze the statistical properties of a polymer chain with pairwise quadratic interactions and algebraically decreasing elastic constants, $A_s \sim s^{-\gamma}$, where $s$ is the chemical distance between monomers and $\gamma \in [2, 4] \setminus \{3\}$. We show that the resulting conformation is Gaussian and self-similar with the fractal dimension, controlled by $\gamma$. Using path integrals, we establish the connection between the quadratic action of a particle with memory-dependent kernel and the partition function of a Gaussian polymer with pairwise interactions.

Introduction

Classical polymer physics describes three primary archetypes: ideal chains (random walks), swollen chains (self-avoiding walks), and collapsed globules. However, complex biological systems, such as mitotic chromosomes or non-concatenated polymer rings, form "topologically stabilized" states known as crumpled globules. These states are compact fractals with a fractal dimension ($d_f$) typically near 3. Modeling these states is challenging due to non-local topological constraints. This work proposes a simplified quadratic Hamiltonian with long-range interactions as an effective model for these complex states.

Theoretical Framework and Methods

The authors employ a Gaussian Network model where monomers are connected by harmonic springs. Unlike standard models, the stiffness of these springs ($a_s$) decays with the chemical distance ($s$) along the chain according to a power law: $a_s \sim s^{-\gamma}$.

Schematic of a polymer chain with pairwise quadratic interactions between monomers.

By treating the network as a Laplacian matrix and performing normal mode analysis, the relaxation modes are determined. Using Euclidean Feynman path integrals, the authors demonstrate that the partition function of this polymer chain is mathematically equivalent to the action of a particle undergoing fractional Brownian motion (fBm). This mapping links the spatial interactions of the polymer to the temporal correlations of a subdiffusive particle through a memory-dependent kernel.

Key Results and Validation

The model establishes a direct relationship between the interaction exponent $\gamma$ and the fractal dimension $d_f$. For the range $2 < \gamma < 3$, the fractal dimension is given by $d_f = 2/(\gamma - 2)$. This allows for a tunable framework where specific fractal dimensions can be achieved by adjusting the interaction decay.

Scaling of the mean square displacement for different interaction regimes.

The model was validated against existing numerical simulation data for unknotted rings (Imakaev et al., 2015) and crumpled globules (Tamm et al., 2015). Despite the topological complexity of these states, monomer-to-monomer distances were found to follow a Gaussian (Maxwell) distribution, justifying the use of a quadratic Hamiltonian. The predicted scaling of the mean square displacement $\sigma^2 \sim s^{2H}$ matches the simulation results, where $H = \gamma/2 - 1$ is the Hurst exponent. For a crumpled globule ($d_f=3$), this corresponds to $H=1/3$.

Discussion and Conclusion

This study reinforces the deep analogy between the static conformation of a polymer in space and the time-trajectory of a diffusing particle. While the model is phenomenological—describing the result of topological constraints rather than their microscopic origin—it provides a powerful tool for describing fractal states.

A significant advantage of this approach is its ability to replicate "crossover" behavior, where a polymer may appear as an ideal random walk at small scales but transitions to a crumpled fractal at larger scales. This multi-scale capability makes the model a strong candidate for describing chromatin packing in the genome, potentially allowing for the mapping of effective interactions from experimental Hi-C data.

Summary updated

Abstract

In the framework of the $\beta$-model proposed by A. Amitai and D. Holcman, we analyze the statistical properties of a polymer chain with pairwise quadratic interactions and algebraically decreasing elastic constants, $A_s \sim s^{-\gamma}$, where $s$ is the chemical distance between monomers and $\gamma \in [2, 4] \setminus \{3\}$. We show that the resulting conformation is Gaussian and self-similar with the fractal dimension, controlled by $\gamma$. Using path integrals, we establish the connection between the quadratic action of a particle with memory-dependent kernel and the partition function of a Gaussian polymer with pairwise interactions.
BibTeX
@misc{polovnikov2017memory,
  author = {Kirill Polovnikov and Sergei Nechaev and M. Tamm},
  title = {{Memory-dependent quadratic action for fractal Brownian motion}},
  year = {2017},
  eprint = {1707.07153},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1707.07153},
}

Loading related content...

Figure