Source-linked AI summary
Coarse-graining DNA for simulations of DNA nanotechnology
Jonathan P. K. Doye, Thomas E. Ouldridge, Ard A. Louis, Flavio Romano, Petr Sulc, Christian Matek, Benedict E. K. Snodin, Lorenzo Rovigatti, John S. Schreck, Ryan M. Harrison, William P. J. Smith
TL;DR
Large-scale, long-timescale DNA nanotechnology processes are inaccessible to all-atom simulation, motivating coarse-grained models that retain essential DNA physics. The paper reviews such approaches and presents oxDNA as a nucleotide-level model suited to nanotechnology and biophysical simulation. oxDNA enables systematic structural and dynamical studies, although complete nanostructure self-assembly remains limited by long timescales and the model has defined salt and structural scope boundaries.
Problem
All-atom simulations cannot probe the large systems and long timescales relevant to DNA nanotechnology, while microscopic mechanisms underlying observed DNA behavior and complex self-assembly remain unclear.
Method
The paper selectively reviews nucleotide-level coarse-grained DNA models and develops oxDNA with rigid, orientationally interacting nucleotides fitted through a top-down approach.
Results
Complete DNA origamis with over ten thousand nucleotides can be structurally characterized, while oxDNA reproduces and provides insight into diverse DNA nanotechnological and biophysical processes.
Takeaways & Limitations
Coarse-grained modelling now allows DNA nanosystems to be systematically and accurately probed, opening DNA nanotechnology to molecular simulations.
Takeaways & Limitations
Direct simulation of large DNA nanostructure self-assembly remains challenging because of the long timescales involved, and oxDNA is parameterized for 500 mM salt.
Abstract
from arXiv · showhide
To simulate long time and length scale processes involving DNA it is necessary to use a coarse-grained description. Here we provide an overview of different approaches to such coarse graining, focussing on those at the nucleotide level that allow the self-assembly processes associated with DNA nanotechnology to be studied. OxDNA, our recently-developed coarse-grained DNA model, is particularly suited to this task, and has opened up this field to systematic study by simulations. We illustrate some of the range of DNA nanotechnology systems to which the model is being applied, as well as the insights it can provide into fundamental biophysical properties of DNA.
1 Introduction
DNA nanotechnology requires models that connect DNA’s biophysical properties to the assembly, operation, and optimization of complex nanostructures. Because all-atom simulations cannot reach relevant scales, the paper motivates coarse-grained models and introduces an overview centered on oxDNA.
- DNA’s base stacking, complementary pairing, and mechanical properties support both biological functions and programmable nanostructure design.
- Complex DNA nanostructures can contain tens of thousands of nucleotides, yet their self-assembly pathways and kinetic traps remain poorly understood.
- Molecular simulations could clarify DNA biophysics, visualize self-assembly and device mechanisms, and support quantitative design optimization.
- All-atom simulations cannot probe the large system sizes and long time scales relevant to DNA nanotechnology.
- Coarse-grained models must retain essential DNA physics while simplifying representation enough for realistic long-scale simulations.
- The paper reviews coarse-grained DNA models, highlights oxDNA, discusses simulation of thermodynamic and dynamic properties, and surveys nanotechnological and biophysical applications.
2 Coarse-grained DNA models: a selective overview
DNA coarse-grained models occupy a spectrum of detail, and models intended for DNA nanotechnology must balance geometric, thermodynamic, mechanical, and computational requirements. The review emphasizes anisotropic interactions and concludes that oxDNA is especially suited to self-assembly studies.
- Models for DNA nanotechnology must represent three-dimensional geometry, hybridization thermodynamics, ssDNA and dsDNA mechanics, and long-time diffusion and rearrangements.
- All-atom and overly simplified models are excluded because they cannot simultaneously provide tractable self-assembly simulation and realistic three-dimensional DNA behavior.
- Coarse-grained models differ in fitting philosophy: bottom-up models fit finer-grained results, whereas top-down models fit targeted experimental or theoretical properties.
- Bottom-up models inherit uncertainties from the fine-grained descriptions, whose reproduction of properties such as DNA melting points may be unknown.
- Interaction potentials commonly include stacking, hydrogen bonding, backbone, excluded-volume, electrostatic, and cross-stacking terms.
- Water coarse-graining makes thermodynamic effects difficult to predict because hydrophobicity can contribute an entropic component to stacking.
- Models should generate dsDNA helicity from combined backbone and stacking constraints so unstacked ssDNA remains flexible and can kink sharply.
- Anisotropic base interactions help preserve coplanar, antiparallel Watson-Crick pairing and avoid nonphysical binding of one base to two complementary bases.
3 The oxDNA model
oxDNA represents DNA as rigid, orientationally interacting nucleotides and uses a top-down parameterization to capture structure, thermodynamics, and mechanics. Its applications span DNA nanostructures and dynamic biophysical processes, while its scope is limited by salt, sequence, and groove assumptions.
- Each oxDNA nucleotide is a rigid body with collinear interaction sites and a perpendicular base vector encoding orientational dependence.
- The model’s orientational interactions represent coplanar stacking, linear hydrogen bonding, edge-to-edge Watson-Crick pairing, and helical strand geometry.
- Coarse-graining each nucleotide as one rigid body removes internal motions and significantly speeds sampling.
- The internucleotide potential combines backbone connectivity, excluded volume, hydrogen bonding, stacking, cross-stacking, and coaxial stacking.
- oxDNA uses top-down parameterization, fitting interaction parameters to structural, thermodynamic, and mechanical data, including duplex melting transitions.
- The model offers average-base and sequence-dependent parameterizations, with the latter varying stacking and hydrogen-bond strengths to capture sequence-dependent thermodynamics.
- oxDNA is parameterized for 500 mM salt, with electrostatic effects incorporated into excluded-volume interactions rather than explicitly represented.
- Its sequence dependence is limited because base properties other than attractive-interaction strengths remain identical, restricting detailed sequence-dependent structure and elasticity.
4 Simulating coarse-grained models
Coarse-grained simulations use specialized dynamics and rare-event methods to recover diffusive motion, improve sampling, and study DNA processes that are difficult to access directly. Method choice depends on system size, target process, and whether equilibrium, structural, or kinetic properties are being measured.
- Diffusive dynamics: Langevin, Brownian, and Andersen-like thermostats produce diffusive motion and sample the canonical ensemble, unlike ballistic dynamics from implicit-solvent Newtonian simulations.Langevin dynamics adds drag and random forces; Andersen-like methods periodically resample momenta.
- Diffusive dynamics: Long-range hydrodynamic forces are usually neglected, so relative diffusion rates of different-sized clusters do not scale correctly.Including hydrodynamic effects is possible but generally avoided because of its extra computational cost.
- Time-scale interpretation: Coarse-grained dynamics should generally be compared through relative rates of similar processes because hydrodynamics and smoothed energy landscapes complicate mapping simulation time to experiment.For oxDNA, artificially accelerated dynamics had no qualitative consequences for duplex formation when relative rates were compared carefully.
- Monte Carlo methods: Cluster-based VMMC avoids the severe inefficiency of single-particle Monte Carlo by constructing moves that reflect both the current configuration and the proposed move.Single-particle moves are often rejected for strongly interacting nucleotides, especially in duplexes and higher-order aggregates.
- Choosing simulation dynamics: VMMC is computationally efficient for small systems and hybridization or melting studies, while Langevin and Andersen-like algorithms are preferred for interpretable kinetics and large-system equilibration.VMMC is also convenient for coupling to umbrella sampling, whereas Andersen-like dynamics efficiently relaxes large structural systems.
- Rare-event methods: Rare-event techniques such as umbrella sampling and parallel tempering improve equilibration across free-energy barriers by enhancing transitions between local minima.Umbrella sampling biases intermediate configurations and corrects the bias to recover an unbiased free-energy landscape.
- Finite-size effects: Single-target simulations of multi-species assembly can suffer finite-size and concentration-fluctuation effects, while bulk estimates require ideal-species and nonaggregating-state assumptions.The authors developed an analytic correction for these effects and recommend applying it to thermodynamic calculations from single-target simulations.
5 Results for oxDNA
OxDNA reproduces key thermodynamic, mechanical, structural, and self-assembly behaviors of DNA while exposing microscopic pathways and kinetic effects relevant to nanotechnology. Its simulations connect molecular mechanisms to experimentally observed responses and design-relevant assembly features.
- Basic DNA biophysics: OxDNA reproduces DNA’s basic double-helix geometry, including pitch, base-pair rise, radius, and emergent propeller twist, but gives both grooves equal size.The propeller twist arises from competition between stacking and hydrogen-bond interactions.
- Basic DNA biophysics: 125 base pairs, 40 bases, and 3 bases are the model’s persistence lengths for dsDNA and stacked and unstacked ssDNA, respectively, broadly matching experiments.The flexibility of unstacked DNA also facilitates hairpin formation and is relevant to nanotechnological applications.
- Basic DNA biophysics: The model captures duplex and hairpin melting behavior, with hairpin melting points reproduced to within an approximately 2 K constant offset.Duplex melting-point dependence on length required balancing stacking and hydrogen-bonding contributions; single strands show a broad uncooperative stacking transition.
- Self-assembly pathways: Hybridization usually proceeds through a nucleus of a few correct base pairs followed by zippering, while misbonded nuclei can resolve through internal displacement.This pathway differs strongly from the mechanism reported for the 3SPN.1 model.
- Self-assembly pathways: Hairpin assembly is fastest within a temperature window: high temperatures hinder complete-stem formation, whereas low temperatures promote misbonded kinetic traps.Preventing non-native base pairing removes the low-temperature decrease in folding rate; annealing exploits the interval between target and misbonded melting temperatures.
- DNA under stress: OxDNA reproduces diverse stress responses, including overstretching at 74 pN, force-induced unpeeling, supercoiling structures, and bending-to-kinking transitions during circular–linear hybridization.The model’s overstretching occurs by unpeeling, while negative supercoiling can produce plectonemes, bubbles, or cruciforms with cooperative but asynchronous arm formation.
6 Conclusions
Coarse-grained DNA models, particularly oxDNA, make systematic simulation of DNA nanosystems feasible across structural, thermodynamic, mechanical, and dynamic processes. OxDNA provides broad physical insight and design guidance, while remaining limited by salt fitting, sequence and pairing simplifications, and four-way-junction chirality.
- 6 Conclusions: OxDNA was specifically designed to capture biophysical processes essential to self-assembling DNA nanotechnology.The model targets structural, thermodynamic, and mechanical properties relevant to DNA nanosystems.
- 6 Conclusions: OxDNA provides physical insight into hybridization, strand exchange, hairpin formation, mechanical stress responses, and DNA nanosystem structures.Its geometric representation also captures nanodevice features unavailable to secondary-structure thermodynamic models, supporting design guidance.
- 6 Conclusions: The main limitations are fitting to 0.5 M salt, limited sequence-dependent elasticity, Watson-Crick-only pairing, a symmetric helix, and incorrect four-way-junction chirality.The model cannot represent alternative DNA forms such as G-quadruplexes and triple-stranded DNA.
- 6 Conclusions: Coarse-grained models now allow DNA nanosystems, including complete DNA origamis with over ten thousand nucleotides, to be systematically and accurately probed.Self-assembly of such large structures remains challenging because of the long time scales involved.