Rigid Base Biasing in Molecular Dynamics enables enhanced sampling of DNA conformations

2022-08-22 · Aderik Voorspoels, Jocelyne Vreede, Enrico Carlon

All-atom simulations have become increasingly popular to study conformational and dynamical properties of nucleic acids as they are accurate and provide high spatial and time resolutions. This high resolution however comes at a heavy computational cost and within the time scales of simulations nucleic acids weakly fluctuate around their ideal structure exploring a limited set of conformations. We introduce the RBB-NA algorithm which is capable of controlling rigid base parameters in all-atom simulations of Nucleic Acids. With suitable biasing potentials this algorithm can "force" a DNA or RNA molecule to assume specific values of the six rotational (tilt, roll, twist, buckle, propeller, opening) and/or the six translational parameters (shift, slide, rise, shear, stretch, stagger). The algorithm enables the use of advanced sampling techniques to probe the structure and dynamics of locally strongly deformed Nucleic Acids. We illustrate its performance showing some examples in which DNA is strongly twisted, bent or locally buckled. In these examples RBB-NA reproduces well the unconstrained simulations data and other known features of DNA mechanics, but it also allows one to explore the anharmonic behavior characterizing the mechanics of nucleic acids in the high deformation regime.

📄 PDF Abstract BibTeX arXiv:2208.10286

Code (1)

aderikvoorspoels/rbb-na 공식 구현

Methods 이 논문이 사용한 방법론

BASE 설명 없음

Kinematic Flexibility Analysis: Hydrogen Bonding Patterns Impart a Spatial Hierarchy of Protein Motion

2018-02-23

Elastic network models (ENM) and constraint-based, topological rigidity analysis are two distinct, coarse-grained approaches to study conformational flexibility of macromolecules. In the two decades since their introduct…

Reinforced dynamics for enhanced sampling in large atomic and molecular systems

2017-12-10 · Linfeng Zhang, Han Wang, Weinan E

A new approach for efficiently exploring the configuration space and computing the free energy of large atomic and molecular systems is proposed, motivated by an analogy with reinforcement learning. There are two major c…

Deep Reinforcement Learning Efficient Exploration reinforcement-learning Reinforcement Learning+1

Unbiasing Enhanced Sampling on a High-dimensional Free Energy Surface with Deep Generative Model

2023-12-14 · YiKai Liu, Tushar K. Ghosh, Guang Lin, Ming Chen

Biased enhanced sampling methods utilizing collective variables (CVs) are powerful tools for sampling conformational ensembles. Due to high intrinsic dimensions, efficiently generating conformational ensembles for comple…

Density Estimation

Tangent Space Least Adaptive Clustering

2021-06-13 · ICML Workshop URL 2021 7 · James Buenfil, Samson J Koelle, Marina Meila

The biasing of dynamical simulations along collective variables uncovered by unsupervised learning has become a standard approach in analysis of molecular systems. However, despite parallels with reinforcement learning (…

Clustering reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

Principal component analysis of nonequilibrium molecular dynamics simulations

2019-05-29

Principal component analysis (PCA) represents a standard approach to identify collective variables $\{x_i\}\!=\!\boldsymbol{x}$, which can be used to construct the free energy landscape $\Delta G(\boldsymbol{x})$ of a mo…