Self-Parametrizing System-Focused Atomistic Models

Christoph Brunken; Markus Reiher

REVIEW 3 major objections 6 minor 162 references

Self-Parametrizing System-Focused Atomistic Models

T0 review · 3 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Atomistic models for arbitrary nanoscale systems can be assembled automatically from one quantum-chemical structure and its Hessian, then refined on the fly by machine learning with uncertainty estimates.

desk verdict A credible proof-of-principle for an integrated SFAM workflow, but the headline accuracy numbers rest on self-sampled configurations and need an external validation before the strong claims hold. read the letter →

arxiv 1908.10492 v2 pith:3M3YPIZM submitted 2019-08-27 physics.chem-ph cond-mat.mes-hallphysics.bio-phphysics.comp-ph

Christoph Brunken , Markus Reiher This is my paper · ORCID

classification physics.chem-phcond-mat.mes-hallphysics.bio-phphysics.comp-ph

keywords self-parametrizingatomisticmodelspartialHessianfittingdeltamachinelearninguncertaintyquantificationmolecularfragmentationmechanicson-the-flyrefinementnanoscalesimulation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper aims to establish that a usable molecular mechanics model for an arbitrary nanoscale system can be built automatically, with no manual parameter fitting, from a single quantum-chemical minimum-energy structure and its Hessian, then refined while it is being used. It combines three pieces: a partial-Hessian fit for the force constants, a fragmentation scheme that makes the parametrization feasible for thousands of atoms, and a Δ-machine-learning correction trained on configurations sampled from the model itself, with cross-validation uncertainties. If the strategy works as claimed, it removes the main bottleneck of atomistic simulation of proteins, surfaces, and nanostructures: the need for hand-tuned force-field parameters for unusual elements or bonding patterns. The authors also argue the same pipeline can generate electrostatic embedding environments for QM/MM models.

What carries the argument

The engine of the workflow is the partial Hessian fitting procedure: each force-constant parameter is obtained by minimizing the Frobenius norm of the difference between the quantum-chemical Hessian submatrix for an atom pair and the corresponding MM Hessian submatrix, with the atom-pair set chosen in each step so that the parameter being fit is the only unknown. The sequential order (dihedral half-barriers first, then angles, then bonds, then impropers) keeps each fit clean of coupling to unoptimized parameters. The other load-bearing object is the Δ-machine-learning correction using a kernel (linear, polynomial, Gaussian, or Laplacian) trained on energies and on atom-centered internal-coordinate forces from the SFAM trajectory; the authors emphasize that the flexibility of the kernel should be matched to the amount of training data.

What would settle it

Run the workflow on a molecule whose SFAM trajectory stays near one conformational basin, then evaluate the hybrid model on configurations from an independent enhanced-sampling simulation that visits other basins, comparing predicted energies and forces to the quantum reference; large errors in those unseen basins would show the refinement loop missed relevant parts of the surface.

Watch

Extended reading notes

Core claim

The central claim is that a "self-parametrizing system-focused atomistic model" (SFAM) can be generated from parameter-free quantum-chemical reference data: equilibrium geometry, Hessian matrix, partial charges, and covalent bond orders. The Hessian is fit in blocks by the partial Hessian method, so each covalent force constant is optimized against only the Hessian submatrices that depend on it; non-covalent terms reuse a dispersion model and a hydrogen-bonding correction with a small set of globally fitted parameters. Large systems are handled by spherical fragments around each atom, whose individual Hessians are assembled into a sparse approximate full Hessian. The resulting MM model is used to run molecular dynamics, and the sampled configurations are sent back for quantum-mechanical reference energies and forces; a kernel ridge regression Δ-machine-learning model then predicts the correction from the MM baseline to the reference, with the standard deviation of 5-fold cross validation used as uncertainty. The paper reports that this hybrid reaches chemical accuracy on energies and forces for butane with far fewer reference calculations than an ML-only model, and that the base MM model alone matches or beats standard force fields on relative energies sampled from an SFAM trajectory.

Load-bearing premise

The accuracy evidence assumes that the configurations visited by the model's own simulation are representative of the configurations that matter; if the potential energy surface is wrong in regions the model never visits, the reported errors and uncertainty estimates are overoptimistic.

Editorial extensions

If this is right

Any molecule whose structure and Hessian can be computed can receive a working MM model with no atom-type library and no manual parameter tuning.
The same model is the trajectory generator, so early inexpensive MD can be used to choose where to spend QM reference calculations, and the model improves as more data arrive.
The system-specific uncertainty from cross validation can be used to decide when new reference data are needed, enabling rolling refinement during a simulation.
The fragmentation parametrization was demonstrated on a 1412-atom protein with 284 fitted force constants, so the scheme is not limited to small molecules; the authors claim it extends to arbitrary nanoscale structures.
The method can be used to build electrostatic MM embedding environments for QM/MM models of reactions, linking the base model to quantum chemistry for bond breaking.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A decisive test the paper does not report would be to sample configurations from an independent, more exploratory ensemble (for example, high-temperature or enhanced-sampling MD) and check whether the uncertainty estimates and errors remain as small as reported; the paper's own comparison is acknowledged to be biased toward SFAM because its test structures came from an SFAM trajectory.
The same self-parametrizing loop could in principle be run with a higher-level reference method on later iterations, yielding a converged hierarchy from cheap semiempirical to coupled-cluster quality PES without changing the MM architecture.
If the base MM model is accurate near equilibrium and the ML correction is localized, then the model's transferability to a new chemical environment should be measurable directly from the growth of the cross-validation uncertainty, giving an operational definition of when reparametrization is needed.
The fragmentation-based redundancy (multiple fragments provide the same atom-pair Hessian blocks) could be exploited not only for missing-data compensation but as a real uncertainty estimate of the force constants themselves, which the paper only suggests.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The paper is a credible integration of known pieces into a single workflow, and it is honest about its limits. The genuinely new thing is the full loop: partial Hessian fitting, automatic fragmentation, MD sampling, delta-ML correction with uncertainty estimates. That loop is absent from the QMDFF/QuickFF literature, and the modular design is clean enough that a reader can see exactly where each piece enters. The partial Hessian fit itself is borrowed from Hirao, and the frequency comparison for 11 molecules (MAE 66.5 cm^-1) is consistent with prior work. The fragmentation study with buffer regions is a useful concrete result: small buffers give errors over 50% in some force constants, larger buffers drop them below roughly 1%. That is a practical finding.

The soft spot is exactly the one the stress-test identifies. The headline energy comparison (Fig. 7) and the whole delta-ML demonstration (Fig. 10) use configurations sampled from an SFAM MD trajectory, not from the DFT ensemble. The paper admits this in Section II.5: the test is biased toward SFAM. That means the reported sub-kcal/mol errors and the cross-validation uncertainty bands show interpolation among SFAM-sampled structures, not accuracy on the target application's configuration distribution. If the base model misses a thermally relevant region, that region never enters the training data, so the rolling refinement loop has no chance to fix it. The plastocyanin section is even weaker: an 800 ps MD run with no reference-energy comparison at all. This is a real limitation of the evidence, not a manufactured one.

I would not call it fatal. The paper is a proof of principle, and the authors flag most of these issues explicitly. But the central claim in the abstract and conclusion, that SFAM provides self-parametrizing system-focused models with on-the-fly improvement, is broader than the demonstrated evidence. The ML proof covers butane in the main text and two molecules in the SI. The workflow still has manual steps: the copper-sulfur bond is added by hand, protonation states and local charges are supplied manually, and two proton-transfer artifacts occurred during fragment optimizations. The implementation is promised for the SCINE project but not yet available.

Who is this for? Methods developers working on automated force-field generation or delta-ML corrections. It deserves serious referee time, because the integration is new and the design is mostly sound, and the SI apparently contains additional examples that could be checked. The main thing a referee should push for is an external validation set: configurations from an unbiased sampling scheme, or at least reweighted trajectories, and a comparison of the hybrid model against reference energies in regions the base model was not trained on. Also give the code and data.

Recommendation: send to peer review. The load-bearing concern is real but fixable with a revised validation, and the paper's honest disclosure of its own bias is a point in its favor.

Referee Report

3 major / 6 minor

Summary. The manuscript presents SFAM, a workflow that automatically parametrizes a classical molecular mechanics potential from a single quantum-chemical minimum-energy structure and its Hessian via partial Hessian fitting, and then improves the model by training a Δ-machine-learning correction on energies and forces of configurations sampled from the SFAM potential. A fragmentation protocol is proposed to scale the parametrization to large systems such as plastocyanin, and the method is demonstrated on small molecules, with vibrational frequencies, relative energies along MD trajectories, and dihedral scans compared against a DFT reference. The central claim is that this combination of an automatically generated physical baseline and an on-the-fly ML correction with uncertainty quantification yields accurate, self-parametrizing atomistic models for arbitrary nanoscale systems.

Significance. If the central claim were fully demonstrated, this would be a valuable step toward removing manual force-field parametrization as a bottleneck in multiscale modeling: the Hessian-based parametrization is well grounded in earlier partial-Hessian fitting work, the fragmentation idea is practical, the integration with Δ-ML and uncertainty estimation is timely, and the authors are transparent about several limitations (e.g., the acknowledged bias of the SFAM-sampled test set). The proof-of-principle on butane and the plastocyanin fragmentation example show that the components work in isolation. However, the validation of the headline accuracy numbers relies on configurations generated by the model being tested, so the current evidence does not establish that the method is accurate on the configuration distribution relevant to a target application, nor does it demonstrate the claimed rolling on-the-fly refinement. The significance is therefore moderate and conditional on additional independent validation.

major comments (3)

[III.3 / Fig. 10 (and II.5 / Fig. 7)] The central accuracy evidence is based on configurations sampled from the SFAM potential itself. In II.5 the authors state that the force-field comparison 'is biased towards SFAM since the structures were sampled from a SFAM trajectory,' and the same bias applies to the Δ-ML results in III.3, where both training and test structures come from a SFAM MD trajectory. Consequently, the reported mean absolute errors below 0.4 kcal/mol and the cross-validation uncertainty bands describe interpolation among configurations that the base model already samples; they do not measure error on the true (DFT) Boltzmann ensemble or on any independent target distribution. Because the paper's central claim is that SFAM provides reliable, uncertainty-quantified models for arbitrary systems, this is a load-bearing issue. I request an additional validation in which the base and Δ-ML models are evaluated on structures generated independently, e.g., from a DFT-based MD trajectory or from a trajectory reweighted to the QM ensemble as the authors themselves suggest.
[III.1 / III.3] The claimed uncertainty quantification is not calibrated. The 5-fold cross-validation standard deviation reported in Fig. 10 and Section III.3 is a measure of model variance on the SFAM-generated training distribution; it is not a prediction interval for the error relative to the reference PES on unseen regions. Section IV states that SFAM provides 'uncertainty quantification that measures the reliability of the model,' but no evidence is shown that the error bars are meaningful, e.g., that the fraction of true errors falling within one or two standard deviations matches the nominal coverage on an independent test set. A calibration study on a held-out independent ensemble is needed to support the uncertainty-quantification component of the central claim.
[III.3 / IV] The rolling, on-the-fly refinement that distinguishes this approach from a one-shot ML fit is not actually demonstrated. In the butane study, a fixed MD trajectory is generated with the base SFAM model, reference data are computed for those configurations, and a single Δ-ML model is trained and evaluated; there is no iteration in which the corrected model generates new configurations, new reference data are added, and the model is improved. The conclusion nevertheless claims that the model 'can be improved in a rolling fashion' and that the approach enables 'on-the-fly' correction. Since this iterative refinement is a central part of the proposed workflow, the manuscript should either include a demonstration of at least one refinement cycle or explicitly restrict the claims to what is shown.

minor comments (6)

[III.1, Eq. (20)] The KRR prediction formula as written, y = (X^T X + λI)^{-1} X^T \tilde{y} x, is dimensionally inconsistent; it should be written with the test vector on the left and the training targets on the right, e.g., y = x_*^T (X^T X + λI)^{-1} X^T \tilde{y}.
[II.4] The plastocyanin example required manual input: the Cu–S bond was added manually because the Mayer bond order was below threshold, and the local charge/spin states were provided manually. The text acknowledges this, but the conclusion's statement that parametrization occurs 'without human interference' is stronger than the demonstrated result; please qualify the claim.
[I / II.1] The phrase 'parameter-free quantum chemical calculations' in the introduction is imprecise, since the reference DFT method and the globally fitted SFAM non-covalent parameters (a1, a2, s8, β, khb) involve fitted or functional parameters. Consider rewording to 'parameters not tailored to the target system' or similar.
[II.2, Eq. (14)] The periodicity rule for dihedral angles is a minor notational issue: the expression (N−1)(M−1)/gcd(N−1, M−1) is the least common multiple of N−1 and M−1; stating this explicitly would aid readability.
[II.5, Fig. 7] The comparison with GAFF, MMFF94, and UFF uses Open Babel implementations of relatively generic force fields; as the authors note, these may not represent the best available specialized force fields. This is acceptable for a proof of principle, but the manuscript should make even clearer that the quantitative comparison is not a benchmark against state-of-the-art specialized parameters.
[III.3] The figure captions for Figs. 10 and 11 state that shaded areas represent confidence intervals, but the text describes them as standard deviations from 5-fold cross-validation. These are not confidence intervals in the statistical sense; please use consistent terminology.

Circularity Check

0 steps flagged · score 2.0 of 10

No constructional circularity; the only self-referential element is the SFAM-sampled validation ensemble, explicitly acknowledged by the authors and not a reduction of any derived quantity to fitted inputs.

full rationale

The derivation chain is not circular. The base MM model's force constants are fitted to an independently calculated quantum-chemical Hessian, and equilibrium values, charges, and bond orders are obtained from first-principles reference calculations, not from the properties later 'predicted.' The global van der Waals and hydrogen-bond parameters are fitted to dimer data and then reused, but the target molecules and their relative energies are not fitted inputs, so those predictions are not forced. The delta-machine-learning corrections target DFT energies and forces, with training and test structures separated by 5-fold cross-validation; the labels are external reference data rather than model outputs. The acknowledged limitation in Section II.5, 'this test is biased towards SFAM since the structures were sampled from a SFAM trajectory,' applies to the force-field comparison and conceptually to the ML validation in Section III.3: the configuration distribution is generated by the SFAM base model, so the reported errors and uncertainty estimates quantify interpolation on the SFAM-sampled ensemble rather than guaranteed accuracy on the full QM equilibrium ensemble. This is a genuine validation weakness and should temper claims of reliability for arbitrary nanoscale targets, but it is not a circular derivation: no equation reduces to its own input, no fitted parameter is renamed as a prediction, and no load-bearing claim rests solely on self-citation. The citations to the authors' prior work on uncertainty quantification (Refs. 55 and 63) are contextual and not used to force the model choice or to forbid alternatives.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The method does not claim parameter-free physics; it assembles a large number of fitted and inherited parameters. The ledger captures the main fitted quantities and domain assumptions. No new physical entities, particles, or forces are introduced.

free parameters (5)

Global dispersion and repulsion parameters a1, a2, s8, beta = a1 = 0.1, a2 = 7.1 bohr, s8 = 4.6, beta = 7.4
Fitted to interaction energies and dissociation curves of 22 molecular dimers in Sec II.3; they set the van der Waals part of the PES for all systems.
Hydrogen-bond element strengths khb = khb(N) = 0.6 a.u., khb(O) = 0.7 a.u., khb(F) = 3.2 a.u., khb(Cl) = 4.2 a.u.
Fitted to interaction energies of small hydrogen-bonded dimers in Sec II.3; controls the E_hb potential.
Per-atom-type covalent force constants = 284 parameters for plastocyanin; values not tabulated in main text
Optimized by partial Hessian fitting in Sec II.3. These are the intended output of the parametrization rather than hidden ad hoc numbers, but they are fitted quantities.
KRR hyperparameters = Not reported; chosen by 5-fold cross-validation
Kernel type, kernel width or degree, and regularization affect the delta-ML correction and are selected on the training data in Sec III.3.
Fragment and atom-type thresholds = fragment radius 5.5 angstrom, minimum fragment size 20 atoms, planarity threshold 20 degrees, Mayer bond-order cutoff…
Chosen by hand in Sec II.2 and II.4; they affect which reference data enter the fit and how many parameters are generated.

assumptions (6)

domain assumption The classical MM functional forms in Eqs. (3)-(13) are a sufficient base approximation to the potential energy surface.
No derivation is given; this is a standard force-field ansatz with harmonic bonds and angles, single-cosine dihedrals, fixed charges, D3 dispersion, exponential repulsion, and QMDFF-style hydrogen bonding.
domain assumption Local partial Hessian blocks contain enough information to determine force constants without fitting to non-equilibrium data.
Adopted from Hirao's partial Hessian fitting in Sec II.3. The paper validates on 11 molecules but does not prove it holds for arbitrary element combinations.
domain assumption D3 atom-pair coefficients and QMDFF effective charges and constants transfer to the SFAM force field.
Section II.1 uses D3 C6, C8 and R0 coefficients and QMDFF Zeff, kappa1 and kappa2 without re-deriving them for the new functional form.
domain assumption Hydrogen-capped fragment Hessians approximate the partial Hessian blocks of the full nanoscale system.
Section II.4 validates this only through force-constant deviations for the OCE molecule, and no full-system energy comparison is made for plastocyanin.
domain assumption PBE-D3/def2-SVP provides sufficiently accurate reference energies and forces for the target applications.
Used throughout as the reference model; errors of the DFT model are not included in the reported uncertainty estimates.
domain assumption Configurations sampled from a base-SFAM MD trajectory cover the configuration space relevant for the target simulation.
This is the sampling basis for both the PES benchmark and the ML training set. The paper acknowledges the benchmark is biased toward SFAM in Sec II.5.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Self-Parametrizing System-Focused Atomistic Models." pith.science (2026). https://pith.science/paper/3M3YPIZM

@misc{pith2026190810492,
  author       = {Pith},
  title        = {Pith review of: Self-Parametrizing System-Focused Atomistic Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3M3YPIZM}},
  note         = {Machine review of arXiv:1908.10492}
}

abstract

Computational studies of chemical reactions in complex environments such as proteins, nanostructures, or on surfaces require accurate and efficient atomistic models applicable to the nanometer scale. In general, an accurate parametrization of the atomistic entities will not be available for arbitrary system classes, but demands a fast automated system-focused parametrization procedure to be quickly applicable, reliable, flexible, and reproducible. Here, we develop and combine an automatically parametrizable quantum chemically derived molecular mechanics model with machine-learned corrections under autonomous uncertainty quantification and refinement. Our approach first generates an accurate, physically motivated model from a minimum energy structure and its corresponding Hessian matrix by a partial Hessian fitting procedure of the force constants. This model is then the starting point to generate a large number of configurations for which additional off-minimum reference data can be evaluated on the fly. A $\Delta$-machine learning model is trained on these data to provide a correction to energies and forces including uncertainty estimates. During the procedure, the flexibility of the machine learning model is tailored to the amount of available training data. The parametrization of large systems is enabled by a fragmentation approach. Due to their modular nature, all model construction steps allow for model improvement in a rolling fashion. Our approach may also be employed for the generation of system-focused electrostatic molecular mechanics embedding environments in a quantum-mechanical/molecular-mechanical hybrid model for arbitrary atomistic structures at the nanoscale.

Figures

Figures reproduced from arXiv: 1908.10492 by the authors.

**Figure 1.** SFAM workflow for generating self-parametrizing system-focused models. An MM model is obtained from an automated fitting procedure to first-principles reference data generated for automatically dissected large molecular structures broken down to reasonably sized fragments (i.e., those for which reference data can be obtained in a time frame small compared to the overall simulation protocol), which can then be employ… view at source ↗

**Figure 2.** Illustration of the internal degrees of freedom contributing to the covalent components of the MM potential energy at the example of ethylmethylamine. We note that some force fields such as AMBER39 and QMDFF64 implement a Fourier series to describe dihedral-angle distortions instead, which leads to a more flexible and, hence, potentially more accurate potential. 91 However, we refrain from such extensions as we inte… view at source ↗

**Figure 3.** Three different partitionings of the pseudo-one-dimensional molecular structure OCE for the subsystem parametrization procedure. The rectangular boxes define the three subsystems (i.e., the line of a box embraces all atoms contributing to the subsystem indicated by that box) for which Hessian matrices are calculated separately. The resulting dangling bonds of these subsystems are saturated by hydrogen atoms. All par… view at source ↗

Figures from the paper (9 more)

**Figure 4.** Figure 4: Overview of the SFAM parametrization procedure for large systems that require fragmentation. The red boxes represent the data objects of the parametrization, while the essential algorithmic subunits of the procedure are presented in blue boxes. The input data is shown …

**Figure 5.** Figure 5: The molecular structures of plastocyanin (left) and its largest fragment with 152 atoms (right) obtained during the SFAM fragmentation procedure. The green arrow points to the central atom of the fragment around which it was constructed. A scheme summarizing the workfl…

**Figure 6.** Figure 6: Comparison of harmonic vibrational frequencies calculated for 11 small organic and inorganic molecules (see Supporting Information) obtained with PBE-D3/def2-SVP and SFAM, respectively. Random colors are assigned to the dots for the sake of clarity. Second, we investig…

**Figure 7.** Figure 7: Comparison of MAE for SFAM relative energies according to Eq. (19) for 10 000 snapshots along a SFAM MD trajectory with three standard force field implementations from the Open Babel toolbox. C2H4ClF and C4H8Cl(OH) refer to the isomers 4-chloro-1- butanol and 1-chloro-…

**Figure 8.** Figure 8: Complete 2π rotational scan around the central bond of butane (red) and 1- chloro-2-fluoroethane (blue) calculated with SFAM and the DFT reference chosen for its parametrization. Both molecules were parametrized at their equilibrium structure, which corresponds to θ = …

**Figure 9.** Figure 9: Complete 2π rotational scan around the peptide bond in the dipeptide dialanine (alanyl-alanine) calculated with SFAM (solid line) and the DFT reference chosen for its parametrization (dashed line). The energies of the equilibrium structures (θ = −180◦ ) were set to zer…

**Figure 10.** Figure 10: Evaluation of the MM/ML hybrid model for electronic energies (top) and forces (bottom) of butane configurations by 5-fold cross validation for differently sized data sets. Each data point corresponds to a molecular structure for which a reference calculation is availa…

**Figure 11.** Figure 11: Evaluation of an ML-only model for electronic energies and forces of butane configurations by 5-fold cross validation for differently sized data sets. Each data point corresponds to a molecular structure for which a reference calculation is available. The shaded areas…

**Figure 12.** Figure 12: Conceptual view of the SFAM approach. We note that the parametrized models produced with our approach may be collected in a universal centralized database (in analogy to the PDB2 and similar to the QCArchive database 161 of the MolSSI project 162) for future applicati…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

162 extracted references · 76 canonical work pages

[1]

Metal-Organic Frameworks: Applications from Catalysis to Gas Storage; John Wiley & Sons, 2011

Farrusseng, D. Metal-Organic Frameworks: Applications from Catalysis to Gas Storage; John Wiley & Sons, 2011

2011
[2]

C.; Koetzle, T

Bernstein, F. C.; Koetzle, T. F.; Williams, G. J.; Meyer Jr, E. F.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. The Protein Data Bank: a computer-based archival ﬁle for macromolecular structures. J. Mol. Biol. 1977, 112, 535–542

1977
[3]

VMD – Visual Molecular Dynamics

Humphrey, W.; Dalke, A.; Schulten, K. VMD – Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33–38

1996
[4]

M.; Lovell, S

Word, J. M.; Lovell, S. C.; Richardson, J. S.; Richardson, D. C. Asparagine and glutamine: using hydrogen atom contacts in the choice of side-chain amide orientation. J. Mol. Biol. 1999, 285, 1735–1747

1999
[5]

M.; Thiel, W

Senn, H. M.; Thiel, W. QM/MM Methods for Biological Systems. Top. Curr. Chem. 2006, 268, 173–290

2006
[6]

Biomolecular Simulations ; Methods in Molecular Biology; Hu- mana Press, Totowa, NJ, 2013; pp 43–66

Groenhof, G. Biomolecular Simulations ; Methods in Molecular Biology; Hu- mana Press, Totowa, NJ, 2013; pp 43–66

2013
[7]

Theoretical studies of enzymic reactions: Dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme

Warshel, A.; Levitt, M. Theoretical studies of enzymic reactions: Dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme. J. Mol. Biol. 1976, 103, 227–249

1976
[8]

Nobel 2013 Chemistry: Methods for computational chemistry

Thiel, W.; Hummer, G. Nobel 2013 Chemistry: Methods for computational chemistry. Nature 2013, 504, 96

2013

Show all 162 references

[9]

J.; Zhang, J.; Klinman, J

Kulik, H. J.; Zhang, J.; Klinman, J. P.; Martinez, T. J. How large should the QM region be in QM/MM calculations? The case of catechol O- methyltransferase. J. Phys. Chem. B 2016, 120, 11381–11394. 39

2016
[10]

Inﬂuence of coupling and embedding schemes on QM size convergence in QM/MM approaches for the example of a proton transfer in DNA

Roßbach, S.; Ochsenfeld, C. Inﬂuence of coupling and embedding schemes on QM size convergence in QM/MM approaches for the example of a proton transfer in DNA. J. Chem. Theory Comput. 2017, 13, 1102–1107

2017
[11]

Heyden, A.; Lin, H.; Truhlar, D. G. Adaptive partitioning in combined quan- tum mechanical and molecular mechanical calculations of potential energy functions for multiscale simulations. J. Phys. Chem. B 2007, 111, 2231–2241

2007
[12]

E.; Ensing, B.; Sikkema, J.; Visscher, L

Bulo, R. E.; Ensing, B.; Sikkema, J.; Visscher, L. Toward a practical method for adaptive QM/MM simulations. J. Chem. Theory Comput. 2009, 5, 2212– 2221

2009
[13]

W.; Laino, T.; Walker, R

Mones, L.; Jones, A.; G¨ otz, A. W.; Laino, T.; Walker, R. C.; Leimkuhler, B.; Cs´ anyi, G.; Bernstein, N. The adaptive buﬀered force QM/MM method in the CP2K and AMBER software packages. J. Comput. Chem. 2015, 36, 633–648

2015
[14]

Zheng, M.; Waller, M. P. Adaptive quantum mechanics/molecular mechanics methods. WIREs Comput. Mol. Sci. 2016, 6, 369–385

2016
[15]

W.; Wang, C.-H.; Garza, C

Duster, A. W.; Wang, C.-H.; Garza, C. M.; Miller, D. E.; Lin, H. Adaptive quantum/molecular mechanics: What have we learned, where are we, and where do we go from here? WIREs Comput. Mol. Sci. 2017, 7, e1310

2017
[16]

P.; Warshel, A

Muller, R. P.; Warshel, A. Ab initio calculations of free energy barriers for chemical reactions in solution. J. Phys. Chem. 1995, 99, 17516–17524

1995
[17]

Convergence of QM/MM free-energy perturbations based on molecular-mechanics or semiempirical simulations

Heimdal, J.; Ryde, U. Convergence of QM/MM free-energy perturbations based on molecular-mechanics or semiempirical simulations. Phys. Chem. Chem. Phys. 2012, 14, 12592–12604

2012
[18]

J.; Pittock, C.; Tautermann, C

Fox, S. J.; Pittock, C.; Tautermann, C. S.; Fox, T.; Christ, C.; Malcolm, N.; Essex, J. W.; Skylaris, C.-K. Free energies of binding from large-scale ﬁrst- principles quantum mechanical calculations: application to ligand hydration energies. J. Phys. Chem. B 2013, 117, 9478–9485

2013
[19]

Cave-Ayland, C.; Skylaris, C.-K.; Essex, J. W. Direct validation of the single step classical to quantum free energy perturbation. J. Phys. Chem. B 2014, 119, 1017–1025

2014
[20]

A.; S¨ oderhjelm, P.; Ryde, U

Mikulskis, P.; Cioloboc, D.; Andreji´ c, M.; Khare, S.; Brorsson, J.; Gen- heden, S.; Mata, R. A.; S¨ oderhjelm, P.; Ryde, U. Free-energy perturbation and quantum mechanical study of SAMPL4 octa-acid host–guest binding en- ergies. J. Comput.-Aided Mol. Des. 2014, 28, 375–400. 40

2014
[21]

S.; Woodcock, H

Hudson, P. S.; Woodcock, H. L.; Boresch, S. Use of nonequilibrium work methods to compute free energy diﬀerences between molecular mechanical and quantum mechanical representations of molecular systems. J. Phys. Chem. Lett. 2015, 6, 4850–4856

2015
[22]

Binding aﬃnities by alchemical per- turbation using QM/MM with a large QM system and polarizable MM model

Genheden, S.; Ryde, U.; S¨ oderhjelm, P. Binding aﬃnities by alchemical per- turbation using QM/MM with a large QM system and polarizable MM model. J. Comput. Chem. 2015, 36, 2114–2124

2015
[23]

S.; Woods, C.; Skylaris, C.-K

Sampson, C.; Fox, T.; Tautermann, C. S.; Woods, C.; Skylaris, C.-K. A step- ping stone approach for obtaining quantum free energies of hydration.J. Phys. Chem. B 2015, 119, 7030–7040

2015
[24]

How many conformations need to be sampled to obtain converged QM/MM energies? The curse of exponential averaging

Ryde, U. How many conformations need to be sampled to obtain converged QM/MM energies? The curse of exponential averaging. J. Chem. Theory Comput. 2017, 13, 5745–5752

2017
[25]

L.; Hudson, P

Kearns, F. L.; Hudson, P. S.; Woodcock, H. L.; Boresch, S. Computing con- verged free energy diﬀerences between levels of theory via nonequilibrium work methods: Challenges and opportunities. J. Comput. Chem. 2017, 38, 1376– 1388

2017
[26]

A.; Ryde, U

Olsson, M. A.; Ryde, U. Comparison of QM/MM methods to obtain ligand- binding free energies. J. Chem. Theory Comput. 2017, 13, 2245–2253

2017
[27]

Boresch, S.; Woodcock, H. L. Convergence of single-step free energy pertur- bation. Mol. Phys. 2017, 115, 1200–1213

2017
[28]

A.; Ignjatovi´ c, M

Caldararu, O.; Olsson, M. A.; Ignjatovi´ c, M. M.; Wang, M.; Ryde, U. Binding free energies in the SAMPL6 octa-acid host–guest challenge calculated with MM and QM methods. J. Comput.-Aided Mol. Des. 2018, 32, 1027–1046

2018
[29]

S.; Boresch, S.; Rogers, D

Hudson, P. S.; Boresch, S.; Rogers, D. M.; Woodcock, H. L. Accelerat- ing QM/MM free energy computations via intramolecular force matching. J. Chem. Theory Comput. 2018, 14, 6327–6335

2018
[30]

R.; Thiel, W.; York, D

K¨ onig, G.; Brooks, B. R.; Thiel, W.; York, D. M. On the convergence of multi-scale free energy simulations. Mol. Simul. 2018, 44, 1062–1081

2018
[31]

D.; Brooks, B.; York, D

K¨ onig, G.; Pickard, F.; Huang, J.; Thiel, W.; MacKerell Jr, A. D.; Brooks, B.; York, D. A comparison of QM/MM simulations with and without the Drude oscillator model based on hydration free energies of simple solutes. Molecules 2018, 23, 2695. 41

2018
[32]

Host–Guest Relative Binding Aﬃnities at Density-Functional Theory Level from Semiempirical Molecular Dynamics Simulations

Wang, M.; Mei, Y.; Ryde, U. Host–Guest Relative Binding Aﬃnities at Density-Functional Theory Level from Semiempirical Molecular Dynamics Simulations. J. Chem. Theory Comput. 2019, 15, 2659–2671

2019
[33]

Mackerell, A. D. Empirical force ﬁelds for biological macromolecules: Overview and issues. J. Comput. Chem. 2004, 25, 1584–1604

2004
[34]

W.; Case, D

Ponder, J. W.; Case, D. A. Advances in Protein Chemistry ; Protein Simula- tions; Academic Press, 2003; Vol. 66; pp 27–85

2003
[35]

Scott, W. R. P.; H¨ unenberger, P. H.; Tironi, I. G.; Mark, A. E.; Billeter, S. R.; Fennen, J.; Torda, A. E.; Huber, T.; Kr¨ uger, P.; van Gunsteren, W. F. The GROMOS Biomolecular Simulation Program Package. J. Phys. Chem. A 1999, 103, 3596–3607

1999
[36]

D.; Daura, X.; Van Gunsteren, W

Schuler, L. D.; Daura, X.; Van Gunsteren, W. F. An improved GROMOS96 force ﬁeld for aliphatic hydrocarbons in the condensed phase. J. Comput. Chem. 2001, 22, 1205–1218

2001
[37]

E.; Van Gunsteren, W

Oostenbrink, C.; Villa, A.; Mark, A. E.; Van Gunsteren, W. F. A biomolecular force ﬁeld based on the free enthalpy of hydration and solvation: the GROMOS force-ﬁeld parameter sets 53A5 and 53A6. J. Comput. Chem. 2004, 25, 1656– 1676

2004
[38]

P.; Choutko, A.; Riniker, S.; Winger, M.; Mark, A

Schmid, N.; Eichenberger, A. P.; Choutko, A.; Riniker, S.; Winger, M.; Mark, A. E.; van Gunsteren, W. F. Deﬁnition and testing of the GROMOS force-ﬁeld versions 54A7 and 54B7. Eur. Biophys. J. 2011, 40, 843

2011
[39]

D.; Cieplak, P.; Bayly, C

Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Fergu- son, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A second generation force ﬁeld for the simulation of proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 1995, 117...

1995
[40]

M.; Caldwell, J

Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Develop- ment and testing of a general amber force ﬁeld. J. Comput. Chem. 2004, 25, 1157–1174

2004
[41]

R.; Bruccoleri, R

Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.; Swami- nathan, S.; Karplus, M. CHARMM: A program for macromolecular energy, minimization, and dynamics calculations. J. Comput. Chem. 1983, 4, 187– 217

1983
[42]

D.; Bashford, D.; Bellott, M.; Dunbrack, R

MacKerell, A. D.; Bashford, D.; Bellott, M.; Dunbrack, R. L.; Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S.; Joseph-McCarthy, D.; Kuchnir, L.; Kuczera, K.; Lau, F. T. K.; Mattos, C.; Michnick, S.; Ngo, T.; 42 Nguyen, D. T.; Prodhom, B.; Reiher, W. E.; Ro...

1998
[43]

Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I.; MacKerell, A. D. CHARMM general force ﬁeld: A force ﬁeld for drug-like molecules compati- ble with the CHARMM all-atom additive biological force ﬁeld...

2010
[44]

W.; MacKerell Jr, A

Pastor, R. W.; MacKerell Jr, A. D. Development of the CHARMM force ﬁeld for lipids. J. Phys. Chem. Lett. 2011, 2, 1526–1532

2011
[45]

Huang, J.; MacKerell Jr, A. D. CHARMM36 all-atom additive protein force ﬁeld: Validation based on comparison to NMR data. J. Comput. Chem. 2013, 34, 2135–2145

2013
[46]

L.; Tirado-Rives, J

Jorgensen, W. L.; Tirado-Rives, J. The OPLS potential functions for proteins, energy minimizations for crystals of cyclic peptides and crambin.J. Am. Chem. Soc. 1988, 110, 1657–1666

1988
[47]

Halgren, T. A. Merck molecular force ﬁeld. I. Basis, form, scope, parameteri- zation, and performance of MMFF94. J. Comput. Chem. 1996, 17, 490–519

1996
[48]

Mapping the genome of meta-generalized gradient approximation density functionals: The search for B97M-V

Mardirossian, N.; Head-Gordon, M. Mapping the genome of meta-generalized gradient approximation density functionals: The search for B97M-V. J. Chem. Phys. 2015, 142, 074111

2015
[49]

Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals

Mardirossian, N.; Head-Gordon, M. Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals. Mol. Phys. 2017, 115, 2315–2372

2017
[50]

J.; Kaasbjerg, K.; Frederiksen, S

Mortensen, J. J.; Kaasbjerg, K.; Frederiksen, S. L.; Nørskov, J. K.; Sethna, J. P.; Jacobsen, K. W. Bayesian error estimation in density-functional theory. Phys. Rev. Lett. 2005, 95, 216401

2005
[51]

T.; Møgelhøj, A.; Petzold, V.; Landis, D

Wellendorﬀ, J.; Lundgaard, K. T.; Møgelhøj, A.; Petzold, V.; Landis, D. D.; Nørskov, J. K.; Bligaard, T.; Jacobsen, K. W. Density functionals for surface science: Exchange-correlation model development with Bayesian error estima- tion. Phys. Rev. B 2012, 85, 235149

2012
[52]

T.; Jacobsen, K

Wellendorﬀ, J.; Lundgaard, K. T.; Jacobsen, K. W.; Bligaard, T. mBEEF: An accurate semi-local Bayesian error estimation density functional. J. Chem. Phys. 2014, 140, 144107. 43

2014
[53]

Prediction uncertainty of density functional approximations for properties of crystals with cubic symmetry

Pernot, P.; Civalleri, B.; Presti, D.; Savin, A. Prediction uncertainty of density functional approximations for properties of crystals with cubic symmetry. J. Phys. Chem. C 2015, 119, 5288–5304

2015
[54]

N.; Reiher, M

Simm, G. N.; Reiher, M. Systematic error estimation for chemical reaction energies. J. Chem. Theory Comput. 2016, 12, 2762–2773

2016
[55]

N.; Reiher, M

Simm, G. N.; Reiher, M. Error-Controlled Exploration of Chemical Reaction Networks with Gaussian Processes. J. Chem. Theory Comput. 2018, 14, 5238– 5248

2018
[56]

Probabilistic performance estimators for computational chemistry methods: The empirical cumulative distribution function of absolute errors

Pernot, P.; Savin, A. Probabilistic performance estimators for computational chemistry methods: The empirical cumulative distribution function of absolute errors. J. Chem. Phys. 2018, 148, 241707

2018
[57]

Grimme, S.; Bannwarth, C.; Shushkov, P. A robust and accurate tight-binding quantum chemical method for structures, vibrational frequencies, and nonco- valent interactions of large molecular systems parametrized for all spd-block elements (Z= 1–86). J. Chem. Theory Comput. 201...

2017
[58]

A Robust Non-Self- Consistent Tight-Binding Quantum Chemistry Method for large Molecules

Pracht, P.; Caldeweyher, E.; Ehlert, S.; Grimme, S. A Robust Non-Self- Consistent Tight-Binding Quantum Chemistry Method for large Molecules. 2019, ChemRxiv, DOI: 10.26434/chemrxiv.8326202.v1

2019 doi
[59]

Calculations of molecules, clusters, and solids with a simpliﬁed LCAO-DFT-LDA scheme

Seifert, G.; Porezag, D.; Frauenheim, T. Calculations of molecules, clusters, and solids with a simpliﬁed LCAO-DFT-LDA scheme. Int. J. Quantum Chem. 1996, 58, 185–192

1996
[60]

N.; Proppe, J.; Reiher, M

Simm, G. N.; Proppe, J.; Reiher, M. Error assessment of computational models in chemistry. Chimia 2017, 71, 202–208

2017
[61]

Reliable Estimation of Prediction Uncertainty for Physicochemical Property Models

Proppe, J.; Reiher, M. Reliable Estimation of Prediction Uncertainty for Physicochemical Property Models. J. Chem. Theory Comput. 2017, 13, 3297– 3317

2017
[62]

Statistical Analysis of Semiclassical Dispersion Corrections

Weymuth, T.; Proppe, J.; Reiher, M. Statistical Analysis of Semiclassical Dispersion Corrections. J. Chem. Theory Comput. 2018, 14, 2480–2494

2018
[63]

Gaussian Process-Based Reﬁnement of Dis- persion Corrections

Proppe, J.; Gugler, S.; Reiher, M. Gaussian Process-Based Reﬁnement of Dis- persion Corrections. J. Chem. Theory Comput. 2019, 15, 6046–6060

2019
[64]

A General Quantum Mechanically Derived Force Field (QMDFF) for Molecules and Condensed Phase Simulations

Grimme, S. A General Quantum Mechanically Derived Force Field (QMDFF) for Molecules and Condensed Phase Simulations. J. Chem. Theory Comput. 2014, 10, 4497–4514. 44

2014
[65]

Reactive force ﬁelds made simple

Hartke, B.; Grimme, S. Reactive force ﬁelds made simple. Phys. Chem. Chem. Phys. 2015, 17, 16715–16718

2015
[66]

QuickFF: A program for a quick and easy derivation of force ﬁelds for metal-organic frameworks from ab initio input

Vanduyfhuys, L.; Vandenbrande, S.; Verstraelen, T.; Schmid, R.; Waro- quier, M.; Van Speybroeck, V. QuickFF: A program for a quick and easy derivation of force ﬁelds for metal-organic frameworks from ab initio input. J. Comput. Chem. 2015, 36, 1015–1027

2015
[67]

R.; Welsh, I

Allison, J. R.; Welsh, I. D. CherryPicker: An Algorithm for the Automated Pa- rameterisation of Large Biomolecules for Molecular Simulation. Front. Chem. 2019, 7, 400

2019
[68]

Comprehensive analysis of the neglect of diatomic diﬀer- ential overlap approximation

Husch, T.; Reiher, M. Comprehensive analysis of the neglect of diatomic diﬀer- ential overlap approximation. J. Chem. Theory Comput. 2018, 14, 5169–5179

2018
[69]

Machine learning force ﬁelds: Construction, validation, and outlook

Botu, V.; Batra, R.; Chapman, J.; Ramprasad, R. Machine learning force ﬁelds: Construction, validation, and outlook. J. Phys. Chem. C 2016, 121, 511–522

2016
[70]

C.; Narayanan, B.; Sen, F

Li, Y.; Li, H.; Pickard IV, F. C.; Narayanan, B.; Sen, F. G.; Chan, M. K.; Sankaranarayanan, S. K.; Brooks, B. R.; Roux, B. Machine learning force ﬁeld parameters from ab initio data. J. Chem. Theory Comput. 2017, 13, 4492– 4503

2017
[71]

Machine learning in materials informatics: recent applications and prospects

Ramprasad, R.; Batra, R.; Pilania, G.; Mannodi-Kanakkithodi, A.; Kim, C. Machine learning in materials informatics: recent applications and prospects. npj Comput. Mater. 2017, 3, 54

2017
[72]

Data driven inference for the repulsive exponent of the Lennard-Jones potential in molecular dynamics simulations

Kulakova, L.; Arampatzis, G.; Angelikopoulos, P.; Hadjidoukas, P.; Papadim- itriou, C.; Koumoutsakos, P. Data driven inference for the repulsive exponent of the Lennard-Jones potential in molecular dynamics simulations. Sci. Rep. 2017, 7, 16576

2017
[73]

P.; De, S.; Poelking, C.; Bernstein, N.; Kermode, J

Bart´ ok, A. P.; De, S.; Poelking, C.; Bernstein, N.; Kermode, J. R.; Cs´ anyi, G.; Ceriotti, M. Machine learning uniﬁes the modeling of materials and molecules. Sci. Adv. 2017, 3, e1701816

2017
[74]

L.; Cs´ anyi, G

Deringer, V. L.; Cs´ anyi, G. Machine learning based interatomic potential for amorphous carbon. Phys. Rev. B 2017, 95, 094203

2017
[75]

P.; Kermode, J.; Bernstein, N.; Cs´ anyi, G

Bart´ ok, A. P.; Kermode, J.; Bernstein, N.; Cs´ anyi, G. Machine learning a general-purpose interatomic potential for silicon. Phys. Rev. X 2018, 8, 041048. 45

2018
[76]

Nature 2019, 566, 195

Reichstein, M.; Camps-Valls, G.; Stevens, B.; Jung, M.; Denzler, J.; Carval- hais, N.; Prabhat, Deep learning and process understanding for data-driven Earth system science. Nature 2019, 566, 195

2019
[77]

A.; Bennie, S

Amabilino, S.; Bratholm, L. A.; Bennie, S. J.; Vaucher, A. C.; Reiher, M.; Glowacki, D. R. Training neural nets to learn reactive potential energy surfaces using interactive quantum chemistry in virtual reality.J. Phys. Chem. A 2019, 123, 4486–4499

2019
[78]

O.; Rupp, M.; von Lilienfeld, O

Ramakrishnan, R.; Dral, P. O.; Rupp, M.; von Lilienfeld, O. A. Big data meets quantum chemistry approximations: The ∆-machine learning approach. J. Chem. Theory Comput. 2015, 11, 2087–2096

2015
[79]

C.; Reiher, M

Husch, T.; Vaucher, A. C.; Reiher, M. Semiempirical molecular orbital models based on the neglect of diatomic diﬀerential overlap approximation. Int. J. Quantum Chem. 2018, 118, e25799

2018
[80]

Explicitly correlated local coupled-cluster methods us- ing pair natural orbitals

Ma, Q.; Werner, H.-J. Explicitly correlated local coupled-cluster methods us- ing pair natural orbitals. WIREs Comput. Mol. Sci. 2018, 8, e1371

2018
[81]

The Density Matrix Renormalization Group in chem- istry and molecular physics: recent developments and new challenges

Baiardi, A.; Reiher, M. The Density Matrix Renormalization Group in chem- istry and molecular physics: recent developments and new challenges. 2019, arXiv preprint arXiv:1910.00137

2019 arXiv
[82]

E.; Poltavsky, I.; Sch¨ utt, K

Chmiela, S.; Tkatchenko, A.; Sauceda, H. E.; Poltavsky, I.; Sch¨ utt, K. T.; M¨ uller, K.-R. Machine learning of accurate energy-conserving molecular force ﬁelds. Sci. Adv. 2017, 3, e1603015

2017
[83]

E.; M¨ uller, K.-R.; Tkatchenko, A

Chmiela, S.; Sauceda, H. E.; M¨ uller, K.-R.; Tkatchenko, A. Towards exact molecular dynamics simulations with machine-learned force ﬁelds. Nat. Com- mun. 2018, 9, 3887

2018
[84]

http://www.rdkit.org/, visited on 2019-05-05

Gregory Landrum, RDKit 2019.03.1. http://www.rdkit.org/, visited on 2019-05-05

2019
[85]

Riniker, S.; Landrum, G. A. Better informed distance geometry: using what we know to improve conformation generation. J. Chem. Inf. Model. 2015, 55, 2562–2574

2015
[86]

N.; Reiher, M

Simm, G. N.; Reiher, M. Context-Driven Exploration of Complex Chemical Reaction Networks. J. Chem. Theory Comput. 2017, 13, 6108–6119

2017
[87]

N.; Vaucher, A

Simm, G. N.; Vaucher, A. C.; Reiher, M. Exploration of Reaction Pathways and Chemical Transformation Networks. J. Phys. Chem. A 2018, 123, 385– 399. 46

2018
[88]

P.; Reiher, M

Unsleber, J. P.; Reiher, M. The Exploration of Chemical Reaction Networks. 2019, arXiv preprint arXiv:1906.10223

2019 arXiv
[89]

Learn on the ﬂy: a hybrid classical and quantum-mechanical molecular dynamics simulation

Cs´ anyi, G.; Albaret, T.; Payne, M.; De Vita, A. Learn on the ﬂy: a hybrid classical and quantum-mechanical molecular dynamics simulation. Phys. Rev. Lett. 2004, 93, 175503

2004
[90]

Harmonic Force Constants for Molecular Me- chanics Force Fields via Hessian Matrix Projection

Allen, A.; Payne, M.; Cole, D. Harmonic Force Constants for Molecular Me- chanics Force Fields via Hessian Matrix Projection. J. Chem. Theory Comput. 2018, 14, 274–281

2018
[91]

Quantum Derivative Fitting and Biomolecular Force Fields: Func- tional Form, Coupling Terms, Charge Flux, Nonbond Anharmonicity, and Individual Dihedral Potentials

Hagler, A. Quantum Derivative Fitting and Biomolecular Force Fields: Func- tional Form, Coupling Terms, Charge Flux, Nonbond Anharmonicity, and Individual Dihedral Potentials. J. Chem. Theory Comput. 2015, 11, 5555– 5572

2015
[92]

Fixed-charge atomistic force ﬁelds for molecular dynamics simula- tions in the condensed phase: An overview

Riniker, S. Fixed-charge atomistic force ﬁelds for molecular dynamics simula- tions in the condensed phase: An overview. J. Chem. Inf. Model. 2018, 58, 565–578

2018
[93]

Charge, bond order and valence in the AB initio SCF theory

Mayer, I. Charge, bond order and valence in the AB initio SCF theory. Chem. Phys. Lett. 1983, 97, 270–274

1983
[94]

On bond orders and valences in the ab initio quantum chemical theory

Mayer, I. On bond orders and valences in the ab initio quantum chemical theory. Int. J. Quantum Chem. 1986, 29, 73–84

1986
[95]

C.; Dasgupta, S.; Lorant, F.; Goddard, W

Van Duin, A. C.; Dasgupta, S.; Lorant, F.; Goddard, W. A. ReaxFF: a reactive force ﬁeld for hydrocarbons. J. Phys. Chem. A 2001, 105, 9396–9409

2001
[96]

P.; Hong, S.; Islam, M

Senftle, T. P.; Hong, S.; Islam, M. M.; Kylasa, S. B.; Zheng, Y.; Shin, Y. K.; Junkermeier, C.; Engel-Herbert, R.; Janik, M. J.; Aktulga, H. M.; Verstrae- len, T.; Grama, A.; van Duin, A. C. T. The ReaxFF reactive force-ﬁeld: de- velopment, applications and future directions. ...

2016
[97]

D.; Feig, M.; Brooks III, C

MacKerell Jr, A. D.; Feig, M.; Brooks III, C. L. Extending the treatment of backbone energetics in protein force ﬁelds: limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J. Comput. Chem. 2004, 25, ...

2004
[98]

D.; Ren, P.; Ponder, J

Rasmussen, T. D.; Ren, P.; Ponder, J. W.; Jensen, F. Force ﬁeld modeling of conformational energies: importance of multipole moments and intramolecular polarization. Int. J. Quantum Chem. 2007, 107, 1390–1395. 47

2007
[99]

W.; Ren, P

Shi, Y.; Xia, Z.; Zhang, J.; Best, R.; Wu, C.; Ponder, J. W.; Ren, P. Polarizable atomic multipole-based AMOEBA force ﬁeld for proteins. J. Chem. Theory Comput. 2013, 9, 4046–4063

2013
[100]

R.; Mackie, I

Johnson, E. R.; Mackie, I. D.; DiLabio, G. A. Dispersion interactions in density-functional theory. J. Phys. Org. Chem. 2009, 22, 1127–1135

2009
[101]

Density functional theory with London dispersion corrections

Grimme, S. Density functional theory with London dispersion corrections. WIREs Comput. Mol. Sci. 2011, 1, 211–228

2011
[102]

A Feynman dispersion correction: a proof of principle for MNDO

Kriebel, M.; Weber, K.; Clark, T. A Feynman dispersion correction: a proof of principle for MNDO. J. Mol. Model. 2018, 24, 338

2018
[103]

The general theory of molecular forces

London, F. The general theory of molecular forces. Trans. Faraday Soc. 1937, 33, 8b–26

1937
[104]

A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu

Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010, 132, 154104

2010
[105]

A thorough benchmark of density functional methods for general main group thermochemistry, kinetics, and noncovalent interac- tions

Goerigk, L.; Grimme, S. A thorough benchmark of density functional methods for general main group thermochemistry, kinetics, and noncovalent interac- tions. Phys. Chem. Chem. Phys. 2011, 13, 6670–6688

2011
[106]

Benchmarking density functional meth- ods against the S66 and S66x8 datasets for non-covalent interactions

Goerigk, L.; Kruse, H.; Grimme, S. Benchmarking density functional meth- ods against the S66 and S66x8 datasets for non-covalent interactions. ChemPhysChem 2011, 12, 3421–3433

2011
[107]

Benchmarking of London dispersion-accounting den- sity functional theory methods on very large molecular complexes

Risthaus, T.; Grimme, S. Benchmarking of London dispersion-accounting den- sity functional theory methods on very large molecular complexes. J. Chem. Theory Comput. 2013, 9, 1580–1591

2013
[108]

Eﬀect of the damping function in disper- sion corrected density functional theory

Grimme, S.; Ehrlich, S.; Goerigk, L. Eﬀect of the damping function in disper- sion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456– 1465

2011
[109]

K.; Casewit, C

Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.; Skiﬀ, W. M. UFF, a full periodic table force ﬁeld for molecular mechanics and molecular dynamics simulations. J. Am. Chem. Soc. 1992, 114, 10024–10035

1992
[110]

Accurate interatomic force ﬁelds via machine learning with covariant kernels

Glielmo, A.; Sollich, P.; De Vita, A. Accurate interatomic force ﬁelds via machine learning with covariant kernels. Phys. Rev. B 2017, 95, 214302

2017
[111]

Assessment of atomic charge models for gas-phase computa- tions on polypeptides

Verstraelen, T.; Pauwels, E.; De Proft, F.; Van Speybroeck, V.; Geerlings, P.; Waroquier, M. Assessment of atomic charge models for gas-phase computa- tions on polypeptides. J. Chem. Theory Comput. 2012, 8, 661–676. 48

2012
[112]

Z.; Tirado-Rives, J.; Jorgensen, W

Vilseck, J. Z.; Tirado-Rives, J.; Jorgensen, W. L. Evaluation of CM5 charges for condensed-phase modeling. J. Chem. Theory Comput. 2014, 10, 2802– 2812

2014
[113]

Hirshfeld, F. L. Bonded-atom fragments for describing molecular charge den- sities. Theor. Chim. Acta 1977, 44, 129–138

1977
[114]

W.; Carb´ o-Dorca, R

Bultinck, P.; Van Alsenoy, C.; Ayers, P. W.; Carb´ o-Dorca, R. Critical analysis and extension of the Hirshfeld atoms in molecules. J. Chem. Phys. 2007, 126, 144111

2007
[115]

V.; Jerome, S

Marenich, A. V.; Jerome, S. V.; Cramer, C. J.; Truhlar, D. G. Charge Model 5: An Extension of Hirshfeld Population Analysis for the Accurate Description of Molecular Interactions in Gaseous and Condensed Phases. J. Chem. Theory Comput. 2012, 8, 527–541

2012
[116]

J.; Trucks, G

Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Petersson, G. A.; Nakatsuji, H.; Li, X.; Caricato, M.; Marenich, A. V.; Bloino, J.; Janesko, B. G.; Gom- perts, R.; Mennucci, B.; Hratchian, H. P.; Ortiz, J...

2009
[117]

C.; Pickard IV, F

Mei, Y.; Simmonett, A. C.; Pickard IV, F. C.; DiStasio Jr, R. A.; Brooks, B. R.; Shao, Y. Numerical study on the partitioning of the molecular polarizabil- ity into ﬂuctuating charge and induced atomic dipole contributions. J. Phys. Chem. A 2015, 119, 5865–5882

2015
[118]

L.; Verstraelen, T.; Matamala, A

Riquelme, M.; Lara, A.; Mobley, D. L.; Verstraelen, T.; Matamala, A. R.; Vohringer-Martinez, E. Hydration Free Energies in the FreeSolv Database Cal- culated with Polarized Iterative Hirshfeld Charges.J. Chem. Inf. Model. 2018, 58, 1779–1797

2018
[119]

N.; T¨ urtscher, P

Simm, G. N.; T¨ urtscher, P. L.; Reiher, M. Systematic Microsolvation Approach 49 with a Cluster-Continuum Scheme and Conformational Sampling.2019, arXiv preprint arXiv:1909.06664

2019 arXiv
[120]

Partial hessian ﬁtting for determining force constant parameters in molecular mechanics

Wang, R.; Ozhgibesov, M.; Hirao, H. Partial hessian ﬁtting for determining force constant parameters in molecular mechanics. J. Comput. Chem. 2016, 37, 2349–2359

2016
[121]

Analytical hessian ﬁtting schemes for eﬃcient determination of force-constant parameters in molecular mechanics

Wang, R.; Ozhgibesov, M.; Hirao, H. Analytical hessian ﬁtting schemes for eﬃcient determination of force-constant parameters in molecular mechanics. J. Comput. Chem. 2018, 39, 307–318

2018
[122]

A method for the solution of certain non-linear problems in least squares

Levenberg, K. A method for the solution of certain non-linear problems in least squares. Q. Appl. Math. 1944, 2, 164–168

1944
[123]

Marquardt, D. W. An algorithm for least-squares estimation of nonlinear pa- rameters. J. Soc. Ind. Appl. Math. 1963, 11, 431–441

1963
[124]

K¨ onig, G.; Brooks, B. R. Correcting for the free energy costs of bond or angle constraints in molecular dynamics simulations. Biochim. Biophys. Acta Gen. Subj. 2015, 1850, 932–943

2015
[125]

S.; Woodcock, H

Hudson, P. S.; Woodcock, H. L.; Boresch, S. On the use of interaction energies in QM/MM free energy simulations. J. Chem. Theory Comput. 2019, 15, 4632–4645

2019
[126]

R.; Stella, M.; Goodpaster, J

Manby, F. R.; Stella, M.; Goodpaster, J. D.; Miller III, T. F. A simple, exact density-functional-theory embedding scheme.J. Chem. Theory Comput. 2012, 8, 2564–2568

2012
[127]

E.; Lee, J.; Miyamoto, K.; Manby, F

Fornace, M. E.; Lee, J.; Miyamoto, K.; Manby, F. R.; Miller III, T. F. Em- bedded mean-ﬁeld theory. J. Chem. Theory Comput. 2015, 11, 568–580

2015
[128]

H.; Reiher, M

M¨ uhlbach, A. H.; Reiher, M. Quantum system partitioning at the single- particle level. J. Chem. Phys. 2018, 149, 184104

2018
[129]

J.; Welborn, M.; Manby, F

Lee, S. J.; Welborn, M.; Manby, F. R.; Miller III, T. F. Projection-Based Wavefunction-in-DFT Embedding. Acc. Chem. Res. 2019, 52, 1359–1368

2019
[130]

R.; Yeates, T

Redinbo, M. R.; Yeates, T. O.; Merchant, S. Plastocyanin: structural and functional analysis. J. Bioenerg. Biomembr. 1994, 26, 49–66

1994
[131]

W.; Zhang, J

Zhang, D. W.; Zhang, J. Molecular fractionation with conjugate caps for full quantum mechanical calculation of protein–molecule interaction energy. J. Chem. Phys. 2003, 119, 3599–3605. 50

2003
[132]

He, X.; Zhang, J. Z. The generalized molecular fractionation with conjugate caps/molecular mechanics method for direct calculation of protein energy. J. Chem. Phys. 2006, 124, 184703

2006
[133]

Z.; He, X

Wang, X.; Liu, J.; Zhang, J. Z.; He, X. Electrostatically embedded generalized molecular fractionation with conjugate caps method for full quantum mechan- ical calculation of protein energy. J. Phys. Chem. A 2013, 117, 7149–7161

2013
[134]

www.mongodb.com, visited on 2019-12-12

MongoDB Inc., MongoDB 3.2. www.mongodb.com, visited on 2019-12-12

2019
[135]

Whitten, J. L. Coulombic potential energy integrals and approximations. J. Chem. Phys. 1973, 58, 4496–4501

1973
[136]

I.; Connolly, J.; Sabin, J

Dunlap, B. I.; Connolly, J.; Sabin, J. On some approximations in applications of X α theory. J. Chem. Phys. 1979, 71, 3396–3402

1979
[137]

Integral approximations for LCAO-SCF calculations

Vahtras, O.; Alml¨ of, J.; Feyereisen, M. Integral approximations for LCAO-SCF calculations. Chem. Phys. Lett. 1993, 213, 514–518

1993
[138]

P.; Burke, K.; Ernzerhof, M

Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865–3868

1996
[139]

Balanced basis sets of split valence, triple zeta va- lence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy

Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence, triple zeta va- lence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297–3305

2005
[140]

The ORCA program system

Neese, F. The ORCA program system. WIREs Comput. Mol. Sci. 2012, 2, 73–78

2012
[141]

Software update: the ORCA program system, version 4.0

Neese, F. Software update: the ORCA program system, version 4.0. WIREs Comput. Mol. Sci. 2018, 8, e1327

2018
[142]

D.; Landis, C

Glendening, E. D.; Landis, C. R.; Weinhold, F. Natural bond orbital methods. WIREs Comput. Mol. Sci. 2012, 2, 1–42

2012
[143]

Intrinsic atomic orbitals: An unbiased bridge between quantum theory and chemical concepts

Knizia, G. Intrinsic atomic orbitals: An unbiased bridge between quantum theory and chemical concepts. J. Chem. Theory Comput. 2013, 9, 4834–4843

2013
[144]

Manz, T. A. Introducing DDEC6 atomic population analysis: part 3. Com- prehensive method to compute bond orders. RSC Adv. 2017, 7, 45552–45581

2017
[145]

Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. Molecular dynamics with coupling to an external bath. J. Chem. Phys. 1984, 81, 3684–3690. 51

1984
[146]

M.; Banck, M.; James, C

O’Boyle, N. M.; Banck, M.; James, C. A.; Morley, C.; Vandermeersch, T.; Hutchison, G. R. Open Babel: An open chemical toolbox. J. Cheminf. 2011, 3, 33

2011
[147]

Hands-on machine learning with Scikit-Learn and TensorFlow ; O’Reilly Media, Inc., 2017

G´ eron, A. Hands-on machine learning with Scikit-Learn and TensorFlow ; O’Reilly Media, Inc., 2017

2017
[148]

J.; Bach, F

Sch¨ olkopf, B.; Smola, A. J.; Bach, F. Learning with kernels: support vector machines, regularization, optimization, and beyond ; MIT press, 2002

2002
[149]

Rupp, M.; Tkatchenko, A.; M¨ uller, K.-R.; von Lilienfeld, O. A. Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning. Phys. Rev. Lett. 2012, 108, 058301

2012
[150]

Machine learning for quantum mechanics in a nutshell

Rupp, M. Machine learning for quantum mechanics in a nutshell. Int. J. Quan- tum Chem. 2015, 115, 1058–1073

2015
[151]

A.; Tkatchenko, A.; Muller, K.-R

Hansen, K.; Montavon, G.; Biegler, F.; Fazli, S.; Rupp, M.; Scheﬄer, M.; Von Lilienfeld, O. A.; Tkatchenko, A.; Muller, K.-R. Assessment and validation of machine learning methods for predicting molecular atomization energies. J. Chem. Theory Comput. 2013, 9, 3404–3419

2013
[152]

Learning scheme to predict atomic forces and accel- erate materials simulations

Botu, V.; Ramprasad, R. Learning scheme to predict atomic forces and accel- erate materials simulations. Phys. Rev. B 2015, 92, 094306

2015
[153]

Adaptive machine learning framework to accelerate ab initio molecular dynamics

Botu, V.; Ramprasad, R. Adaptive machine learning framework to accelerate ab initio molecular dynamics. Int. J. Quantum Chem. 2015, 115, 1074–1083

2015
[154]

R.; De Vita, A

Li, Z.; Kermode, J. R.; De Vita, A. Molecular dynamics with on-the-ﬂy ma- chine learning of quantum-mechanical forces. Phys. Rev. Lett. 2015, 114, 096405

2015
[155]

Machine learning for atomic forces in a crystalline solid: Transferability to various temperatures

Suzuki, T.; Tamura, R.; Miyazaki, T. Machine learning for atomic forces in a crystalline solid: Transferability to various temperatures. Int. J. Quantum Chem. 2017, 117, 33–39

2017
[156]

Rupp, M.; Ramakrishnan, R.; von Lilienfeld, O. A. Machine learning for quan- tum mechanical properties of atoms in molecules. J. Phys. Chem. Lett. 2015, 6, 3309–3313

2015
[157]

Regularised Atomic Body-Ordered Permutation-Invariant Polynomials for the Construction of In- teratomic Potentials

van der Oord, C.; Dusson, G.; Csanyi, G.; Ortner, C. Regularised Atomic Body-Ordered Permutation-Invariant Polynomials for the Construction of In- teratomic Potentials. 2019, arXiv preprint arXiv:1910.06010

2019 arXiv
[158]

Monte Carlo and molecular dynamics simulations in polymer sci- ence; Oxford University Press, 1995

Binder, K. Monte Carlo and molecular dynamics simulations in polymer sci- ence; Oxford University Press, 1995. 52

1995
[159]

Understanding molecular simulation: from algorithms to applications; Elsevier, 2001; Vol

Frenkel, D.; Smit, B. Understanding molecular simulation: from algorithms to applications; Elsevier, 2001; Vol. 1

2001
[160]

http://scine.ethz

SCINE − Software for Chemical Interaction Networks. http://scine.ethz. ch/, visited on 2019-06-17

2019
[161]

https://qcarchive.molssi.org/, visited on 2019-07-11

The QCArchive developers, QCArchive − The MolSSI Quantum Chemistry Archive. https://qcarchive.molssi.org/, visited on 2019-07-11

2019
[162]

Wilkins-Diehr, N.; Crawford, T. D. NSFs inaugural software institutes: The science gateways community institute and the molecular sciences software in- stitute. Comput. Sci. Eng. 2018, 20, 26–38. 53

2018

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Metal-Organic Frameworks: Applications from Catalysis to Gas Storage; John Wiley & Sons, 2011

Farrusseng, D. Metal-Organic Frameworks: Applications from Catalysis to Gas Storage; John Wiley & Sons, 2011

2011

[2] [2]

C.; Koetzle, T

Bernstein, F. C.; Koetzle, T. F.; Williams, G. J.; Meyer Jr, E. F.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. The Protein Data Bank: a computer-based archival ﬁle for macromolecular structures. J. Mol. Biol. 1977, 112, 535–542

1977

[3] [3]

VMD – Visual Molecular Dynamics

Humphrey, W.; Dalke, A.; Schulten, K. VMD – Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33–38

1996

[4] [4]

M.; Lovell, S

Word, J. M.; Lovell, S. C.; Richardson, J. S.; Richardson, D. C. Asparagine and glutamine: using hydrogen atom contacts in the choice of side-chain amide orientation. J. Mol. Biol. 1999, 285, 1735–1747

1999

[5] [5]

M.; Thiel, W

Senn, H. M.; Thiel, W. QM/MM Methods for Biological Systems. Top. Curr. Chem. 2006, 268, 173–290

2006

[6] [6]

Biomolecular Simulations ; Methods in Molecular Biology; Hu- mana Press, Totowa, NJ, 2013; pp 43–66

Groenhof, G. Biomolecular Simulations ; Methods in Molecular Biology; Hu- mana Press, Totowa, NJ, 2013; pp 43–66

2013

[7] [7]

Theoretical studies of enzymic reactions: Dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme

Warshel, A.; Levitt, M. Theoretical studies of enzymic reactions: Dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme. J. Mol. Biol. 1976, 103, 227–249

1976

[8] [8]

Nobel 2013 Chemistry: Methods for computational chemistry

Thiel, W.; Hummer, G. Nobel 2013 Chemistry: Methods for computational chemistry. Nature 2013, 504, 96

2013