This Title All WIREs
How to cite this WIREs title:
WIREs Comput Mol Sci
Impact Factor: 8.127

# Reactive molecular dynamics: From small molecules to proteins

Can't access this content? Tell your librarian.

The current status of reactive molecular dynamics (MD) simulations is summarized. Both, methodological aspects and applications to problems ranging from gas phase reaction dynamics to ligand‐binding in solvated proteins are discussed, focusing on extracting information from simulations that cannot easily be obtained from experiments alone. One specific example is the structural interpretation of the ligand rebinding time scales extracted from state‐of‐the art time‐resolved experiments. Atomistic simulations employing validated reactive interaction potentials are capable of providing structural information about the time scales involved. Both, merits and shortcomings of the various methods are discussed and the outlook summarizes possible future avenues such as reactive potentials based on machine learning techniques. This article is categorized under: Molecular and Statistical Mechanics > Molecular Dynamics and Monte‐Carlo Methods Theoretical and Physical Chemistry > Reaction Dynamics and Kinetics Molecular and Statistical Mechanics > Molecular Interactions
(a) Schematic illustrating the ARMD method for a collinear reaction, where atom B is transferred from donor atom A to acceptor atom C. During crossing the surfaces are switched in time and the Morse bond is replaced by van der Waals (vdW) interactions and vice versa. b) Simple model for estimating energy violation in ARMD simulations. The system with mass m approaches from the left on potential energy surface VR(x) = αx (phase I). At time t = 0 it is at x0 with velocity v0 and kinetic energy Ekin, 0. After crossing is detected at x = 0 the time is rewound by ts/2 and the dynamics is resimulated while VR(x) is being switched to VP(x) = βx in ts (phase II)
[ Normal View | Magnified View ]
Free energy profile including error bars for the ligand exchange reaction in WT trHbN from umbrella sampling (black trace) along the reaction coordinate . The structures of O2‐bound (ρ = 0.25), NO‐bound (ρ = 2.5), and the transition state (ρ ∼ 1.0) are also shown. The TS region is broad and involves a distributed structural ensemble. The red dashed line shows the free energy profile for the ligand exchange reaction for the Y33A mutant. (Reprinted with permission from Reference )
[ Normal View | Magnified View ]
Left: Time dependence of the Fe‐oop motion after photodissociation of the ligand and subsequent dynamics on the 4A PES. The black line is the fit to and the blue and green lines are the experimentally determined exponential decays.187 Right: NO‐rebinding kinetics depending on the value of the asymptotic separation Δ. (Reprinted with permission from Reference 135. Copyright 2015 American Institute of Physics and Reference 176)
[ Normal View | Magnified View ]
Quality of the reactive force fields for the photodissociation of H2SO4 (left) and HSO3F (right). In the left panel points along the minimum energy paths are red squares and the remaining training points from finite‐temperature molecular dynamics simulations are black circles. In the right panel, the red, black, green, blue, and yellow symbols correspond to sampling the νOH = 6 overtone, HSO3Cl at 300 K, the minimum energy path, the dihedral dynamics, and the HCl + SO3 complex, respectively. (Reprinted with permission from Reference 57. Copyright 2016 American Chemical Society and Reference 174. Copyright 2014 American Chemical Society)
[ Normal View | Magnified View ]
Vibrational energy distribution p(ν) for the N2 product following the NO(ν = j = 0) + NN2 + O reaction with J = 0 from time independent (red lines with circles) and QCT (black line with squares) simulations on both PESs, 3A and 3A′′. (Reprinted with permission from Reference 161. Copyright 2017 American Institute of Physics)
[ Normal View | Magnified View ]
Rate coefficients k(T) (blue open triangles) and k+(T) (red open triangles) for the NO(X2Π) + N(4S) N2(X1Σ) + O(3P) reaction for 300 ≤ T ≤ 20,000 K. The inset shows results for 1,000 ≤ T ≤ 5,000 K. The labels correspond to references: (a) = Reference , (b) = Reference , (c) = Reference ), (d) = Reference ), and (e) = Reference . (Reprinted with permission from Reference 145. Copyright 2015 American Institute of Physics)
[ Normal View | Magnified View ]
Contour plots for the NO1 + O2 (panel a) and O2O1 + N (panel b) grids in R1, α1 and R3, α3 Jacobi coordinates, respectively. The N‐O1 (r1) and O2‐O1 (r3) distances are 2.25 a0 and the contour levels are separated by 0.5 eV. (c) MRCI+Q/cc‐pVQZ energies along the R coordinate for r = 1.21A (N‐O1 distance) and α = 27.5, 34.0, and 43.1° (squares, circles and triangles, respectively). The solid lines are RKHS interpolants. The inset in the graph represents a close‐up of the cut for α = 34.0° which is not used for the RKHS interpolation. (d) Performance of the RKHS method for the function . The RKHS interpolant (red dashed line) constructed from the training samples (black dots) is virtually identical to the analytical expression (gray line). (Reprinted with permission from Reference 35. Copyright 1990 American Chemical Society and Reference 226. Copyright 2006 American Institute of Physics)
[ Normal View | Magnified View ]
The MS‐ARMD switching method applied in one and two dimensions to 3 and 2 surfaces (V1, 2, 3). The effective surface (VMS‐ARMD) is always close to the lowest‐energy surface (Vmin), except for regions where other surfaces are within a few times ΔV (here = 0.5) in energy. Here, the algorithm switches smoothly among them by varying their weights (w1, 2, 3; lower left panel)
[ Normal View | Magnified View ]