Pith. sign in

REVIEW 4 major objections 3 minor 87 references

Training a force field for proteins and small molecules from scratch

T0 review · 4 major / 3 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read Garnet learns a full biomolecular force field from scratch, with no legacy parameters

desk verdict A genuinely from-scratch GNN force field with real benchmarks, but the NMR training gradient discrepancy and a few post hoc exclusions keep it from being fully clean. read the letter →

arxiv 2603.16770 v2 pith:ZMJI74PC submitted 2026-03-17 q-bio.BM

Alexandre Blanco-González , Thea K Schulze , Evianne Rovers , Joe G Greener This is my paper · ORCID
classification q-bio.BM
keywords forcefieldgraphneuralnetworkcontinuousatomtypingmoleculardynamicsdoubleexponentialpotentialensemblereweightingbindingfreeenergyNMRtraining
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

This paper claims that a graph neural network, Garnet, can assign every parameter needed for classical molecular dynamics simulations—bonds, angles, torsions, partial charges, and non-bonded interactions—for arbitrary proteins and small molecules, without reusing any existing force field parameters. The model is trained on quantum mechanical data, condensed-phase enthalpies, and protein NMR observables, using differentiable simulation and ensemble reweighting. The authors argue that the resulting force field performs comparably to established force fields on small-molecule conformational benchmarks, folded proteins, protein complexes, and disordered proteins, and that it gives relative binding free energy predictions similar to widely used protocols. If this holds, it would mean that manual, system-specific force field parameterization can be replaced by an automated, reproducible pipeline, and that new functional forms can be tested systematically rather than being locked in by historical parameter sets.

What carries the argument

The central object is the GNN-based continuous atom typing model, which predicts all force field parameters from bonding topology alone, using atom embeddings with a two-bond receptive field. The load-bearing mechanism is the double exponential potential, which replaces Lennard-Jones for non-bonded interactions and is trainable from scratch, and the ensemble reweighting gradient estimator that allows NMR data to backpropagate into the parameters through simulations. Training also splits DFT forces into intramolecular and intermolecular parts, weights the weak intermolecular signal more heavily, and co-trains water parameters without special treatment.

What would settle it

Run the published training code and verify that the implemented ensemble-reweighting gradient matches the equation in the Methods, then retrain Garnet from scratch without the initial GAFF/Amber14SB conformations and compare its GB3 NMR benchmark metrics; if the gradient is biased or the initial conformations are essential, the central 'from scratch' claim would be undermined.

Watch

Extended reading notes

Core claim

The central discovery is that an end-to-end differentiable pipeline can train all molecular mechanics force field parameters from data, without seeding from existing parameters. Garnet uses a graph neural network whose continuous atom embeddings are converted into bond, angle, torsion, charge, and non-bonded parameters, and these parameters are optimized by running simulations during training and matching quantum mechanical forces and energies, experimental enthalpies, and NMR scalar couplings and chemical shifts. A key technical finding is that the conventional Lennard-Jones potential is unstable to train from scratch with this approach, whereas a double exponential potential—with two globa

Load-bearing premise

The ensemble reweighting gradient estimator used to train on GB3 NMR data works correctly despite the paper's admission that the implemented code differed from the stated equation, and the initial training conformations came from existing force fields, so if the estimator is biased the model's optimization on NMR data is not as described.

Editorial extensions

If this is right

  • If the claims hold, an open, reproducible pipeline can replace manual force field parameterization, enabling systematic exploration of new functional forms and rapid retraining as more data become available.
  • A single parameterization scheme covering proteins, small molecules, and water, with no molecule-specific tuning, would simplify and standardize molecular dynamics simulations across biological and chemical applications.
  • The finding that the double exponential potential is more amenable to automated training than Lennard-Jones suggests that future force field development should reconsider the standard non-bonded functional form.
  • Competitive relative binding free energy predictions with an automatically parameterized force field would lower the barrier to using rigorous alchemical calculations in drug discovery, especially for novel chemical matter.
  • The approach's reliance on experimental protein NMR data during training opens the door to directly optimizing force fields against biologically relevant observables, not just quantum mechanical energies.

Reading between the lines

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

  • The result implicitly challenges the assumption that decades of human tuning are needed for transferable force fields; a purely data-driven starting point may generalize to molecule classes that are underserved by existing parameter sets, such as intrinsically disordered proteins.
  • If the double exponential potential proves robust in other packages and for other molecule types, it may become a standard alternative to Lennard-Jones, and the paper's comparison of functional forms provides a template for benchmarking such changes fairly.
  • The identified weakness in handling net-charge-changing alchemical transformations suggests a concrete target: training force fields with charge-dependent or polarizable terms, or using free energy data directly as a training signal, which the authors note as a future direction.
  • The use of initial trajectories generated by existing force fields means the 'from scratch' claim is partially softened; a fully independent version, trained without any reference to GAFF or Amber14SB conformations, would be a stronger test of the approach.
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.

Referee Report

4 major / 3 minor

Summary. Garnet is a graph neural network that predicts all molecular mechanics parameters—bond, angle, torsion, partial charge, and non-bonded terms—from molecular topology, with no reuse of legacy force-field parameters. It is trained on DFT data (SPICE, GEMS, MACE-OFF), condensed-phase enthalpies of vaporisation/mixing, and GB3 NMR observables via ensemble reweighting, and it uses a double exponential non-bonded potential rather than Lennard-Jones. The authors benchmark Garnet on SPICE forces/energies, OpenFF Industry conformer minimisation, folded-protein and protein-complex simulations, IDP radii of gyration, water properties, and relative binding free energies with OpenFE. They report performance comparable to Amber14SB, Espaloma, and OpenFF, and in the RBFE setting similar to OpenFE and FEP+.

Significance. If the central claim holds, the work is significant: it demonstrates that a complete classical force field can be produced by an automated, data-driven pipeline, and it identifies the double exponential potential as a trainable alternative to Lennard-Jones. The paper's strengths include open code and data, a transparent functional-form comparison, and benchmarks spanning small molecules, proteins, complexes, IDPs, and binding free energies. However, the 'from scratch' claim is currently weakened by an explicit discrepancy between the published ensemble-reweighting equation and the code actually used for NMR training, and by post hoc exclusions in the SPICE and RBFE benchmarks. These issues are local and potentially fixable, but they need to be resolved before the headline claims can be fully accepted.

major comments (4)
  1. [Methods, Neural network training (ensemble reweighting equation)] The text states: 'Due to a mistake noticed after training, the actual code used differed from the above equation as described in the source code.' Since the GB3 NMR data are one of the three training modalities that distinguish this work from prior GNN-based force fields, the paper must specify what objective was actually optimised. If the implemented gradient differs from the stated estimator, the claim that Garnet was trained on protein NMR data is not currently established from the manuscript alone. The printed equation also does not transparently show the normalisation on both terms of the reweighting gradient. Please provide the implemented estimator, a corrected display equation, and a numerical check (e.g., gradient comparison on a small system or a demonstration that the intended loss decreased) so that the training story is verifiable.
  2. [Results, Relative binding free energy benchmark; Supplementary Table 3; Figure 6] The main-text 'Overall' metrics are computed over eight targets, but Table 2 lists nine systems, with cmet relegated to Supplementary Table 3. The cmet series has a known net-charge-change failure, and Garnet's pairwise RMSE there is 8.26 kcal/mol versus 1.99 for OpenFE and 1.07 for FEP+. Excluding this failing target from the headline claim 'similar results to popular approaches for relative binding free energy predictions' is post hoc. Please report overall metrics including cmet, or pre-specify and justify the exclusion criteria. As written, the reader cannot distinguish a targeted benchmark from a favourable subset of the data.
  3. [Table 1 (SPICE benchmark)] The table reports only conformations that were 'successful with all four methods.' If one method fails on a substantial number of conformations, comparison on the common subset can be optimistic or biased in either direction. Please report the number of conformations excluded per method and per subset, and check whether the qualitative conclusions survive when each method is evaluated on the conformations it can handle individually. The current presentation does not allow the reader to assess the impact of this filtering.
  4. [Results, Protein benchmark (Figure 3B-C)] GB3 was used during training and also serves as a folded-protein benchmark. The authors note that the good GB3 performance 'could be due to overfitting', which is appropriate, but this means GB3 is not independent evidence of transferability. The independent proteins (BPTI, HEWL, Ubq) should carry the folded-protein claim, and GB3 should be explicitly labelled as a train-set check in the figure or main text rather than presented as part of the validation set.
minor comments (3)
  1. [Methods, Neural network training] Typo in the proper-torsion regularisation loss: 'paramteri' should be 'parameter i'.
  2. [Methods, Relative binding free energy benchmark] The manuscript uses openfe-v1.8.0 while the OpenFE Industry Benchmarking Project used openfe-v1.0.1. Please state whether this version difference changes the default protocol settings and justify that the comparison still isolates the force field rather than the software version.
  3. [Results, Different functional forms] Figure 5C caption says 'models trained from scratch' for all variants, but the supplementary methods clarify that these runs used only reference trajectories. Please make this distinction in the main text to avoid ambiguity.

Circularity Check

1 steps flagged · score 2.0 of 10

No circular derivation; one mild training/benchmark overlap on GB3, disclosed by the paper.

  1. fitted input called prediction [Results, Protein benchmark (Figure 3); Methods, Neural network training (GB3 NMR ensemble reweighting losses)]
    "We included GB3, which was used during training, to assess possible overtraining by the model. ... The good performance on GB3 by Garnet on these benchmarks could be due to overfitting."

    Garnet's parameters were optimised against GB3 NMR J-couplings and chemical shifts via ensemble reweighting, so GB3 is not an independent test of the protein component. Presenting GB3 among the 'folded proteins' results in Figure 3 and the abstract means part of the headline protein performance is a check of fit rather than an external prediction. The paper itself flags this possibility, so the issue is mild; the other folded-protein benchmarks (Ubq, BPTI, HEWL) are independent and support the broader claim.

full rationale

The central derivation is a parameter-fitting pipeline: QM forces/energies/charges, condensed-phase enthalpies, and protein NMR data are used as training targets, and benchmark properties are then computed from fresh simulations. No step in the paper reduces by construction to its own inputs: the ensemble reweighting equation is a gradient estimator rather than a definition of the target losses, the double-exponential functional form is selected empirically by training stability and validation loss, and the reported small-molecule, IDP, complex, and RBFE results are not algebraic consequences of the fitted parameters. The admitted discrepancy between the published NMR gradient equation and the code actually used is a reproducibility/correctness problem, not a circularity. The early use of GAFF/TIP3P and Amber14SB/TIP3P trajectories is disclosed and affects initialization, not the claimed equivalence of prediction and input. Self-citations to Molly.jl and BioStructures.jl are tooling citations, not load-bearing evidence. The only notable overlap is GB3, which was used in training and then appears in the protein benchmark; because the paper explicitly cautions that good GB3 performance could be overfitting, this is a mild evaluation leak rather than a forced circular result.

Assumptions & free parameters 7 free parameters · 7 assumptions · 0 invented entities

The model's parameters are outputs of a trained GNN, so the effective free parameters are the network weights plus the chosen functional form and global nonbonded parameters. The central claims rest on domain assumptions about DFT data combination, force decomposition, ensemble reweighting correctness, and the benign effect of early reference trajectories from existing force fields.

free parameters (7)
  • global double exponential α and β = not reported
    Global parameters of the nonbonded potential, trained on QM and condensed-phase data (Methods, Functional forms; Fig 5C).
  • per-atom double exponential σ = range [0.05, 0.5] nm
    GNN output via sigmoid; trained on DFT and condensed phase data (Methods, Neural network architecture).
  • per-atom double exponential ε = range [0.02, 1.5] kJ/mol
    GNN output via sigmoid; trained on DFT and condensed phase data.
  • partial charges via electronegativity/hardness = not reported as single values
    Trained to MBIS partial charges from SPICE and charge loss (Methods, Neural network architecture, Neural network training).
  • bond, angle, proper/improper torsion parameters = various
    FCNN outputs for harmonic bonds/angles and cosine torsions, trained on DFT forces and energy differences.
  • 1-4 Coulomb weighting and nonbonded global parameters = not reported
    Trained alongside other parameters; 1-4 weighting set to zero for double exponential due to OpenMM constraints (Methods, Neural network architecture).
  • GNN weights = 207,187 parameters total
    All network weights are fitted to the combined DFT/experimental training data.
assumptions (7)
  • domain assumption Combining DFT data from different functionals (SPICE, GEMS, MACE-OFF, Espaloma) is valid because functional differences are small compared to MM error
    Methods, Neural network training: 'differences due to the functionals are likely to be small compared to the high error inherent in MM functional forms'.
  • domain assumption Splitting DFT forces into intra- and intermolecular contributions by overall translation/rotation correctly identifies intermolecular forces
    Methods, Neural network training: the approach 'considers the overall translation and rotation force on the molecule as arising from intermolecular forces'.
  • ad hoc to paper The implemented ensemble reweighting gradient estimator (which differs from the stated equation) gives a useful training signal for NMR data
    Methods, Neural network training: 'Due to a mistake noticed after training, the actual code used differed from the above equation as described in the source code.'
  • ad hoc to paper Early training trajectories from GAFF/TIP3P and Amber14SB/TIP3P do not bias final parameters beyond providing conformations
    Methods, Neural network training: 'Technically this means that the force field has some dependence on existing force fields, but only via generated trajectories and only early in training.'
  • domain assumption Karplus relationships and Graph NMR chemical shift predictor are accurate enough to train and evaluate protein NMR observables
    Methods, Neural network training lists GB3 J-coupling and chemical shift losses using Karplus parameters and Graph NMR [71].
  • domain assumption The double exponential potential with Lorentz-Berthelot combination rules can capture the nonbonded physics needed for proteins and small molecules
    This functional form was chosen because the Lennard-Jones potential produced unstable training simulations; no independent physical derivation is given.
  • domain assumption Including all atoms, including water hydrogens, in the nonbonded potential is a valid modelling choice
    Methods, Neural network architecture: 'In contrast to water models where the hydrogen atoms do not take part in Lennard-Jones interactions, here all atoms are included.'

how reviews work

0 comments
Cite this review

Pith. "Pith review of Training a force field for proteins and small molecules from scratch." pith.science (2026). https://pith.science/paper/ZMJI74PC

@misc{pith2026260316770,
  author       = {Pith},
  title        = {Pith review of: Training a force field for proteins and small molecules from scratch},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZMJI74PC}},
  note         = {Machine review of arXiv:2603.16770}
}
read the original abstract

Force fields for molecular dynamics are usually developed manually, limiting their transferability and making systematic exploration of functional forms challenging. We developed a graph neural network that assigns all force field parameters for diverse molecules using continuous atom typing. The freely-available model, called Garnet, was trained on quantum mechanical, condensed phase and protein nuclear magnetic resonance data without the use of existing parameters. The resulting force field shows comparable performance to current force fields on small molecules, folded proteins, protein complexes and disordered proteins. It shows similar results to popular approaches for relative binding free energy predictions across a range of targets. Assessing different functional forms shows that the double exponential potential is a flexible and accurate alternative to the Lennard-Jones potential. Garnet provides a platform for automated, reproducible force field discovery that brings the benefits of machine learning to classical force fields.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

87 extracted references · 2 linked inside Pith

  1. [1]

    Fixed-Charge Atomistic Force Fields for Molecular Dynamics Simulations in the Condensed Phase: An Overview,

    S. Riniker, “Fixed-Charge Atomistic Force Fields for Molecular Dynamics Simulations in the Condensed Phase: An Overview,”Journal of Chemical Information and Modeling, vol. 58, no. 3, pp. 565–578, 2018

    2018

  2. [2]

    Recent Developments in Amber Biomolecular Simulations,

    D. A. Case, D. S. Cerutti, V. W. D. Cruzeiro, T. A. Darden, R. E. Duke, M. Ghazimirsaeed, G. M. Giamba¸ su, T. J. Giese, A. W. G¨ otz, J. A. Harris, K. Kasavajhala, T.-S. Lee, Z. Li, C. Lin, J. Liu, Y. Miao, R. Salomon-Ferrrer, J. Shen, R. Snyder, J. Swails, R. C. Walker, J. Wang, X. Wu, J. Zeng, T. E. Cheatham III, D. R. Roe, A. Roitberg, C. Simmerling, ...

    2025

  3. [3]

    CHARMM: The biomolecular simulation program,

    B. R. Brooks, C. L. Brooks III, A. D. Mackerell Jr., L. Nilsson, R. J. Petrella, B. Roux, Y. Won, G. Archontis, C. Bartels, S. Boresch, A. Caflisch, L. Caves, Q. Cui, A. R. Dinner, M. Feig, S. Fischer, J. Gao, M. Hodoscek, W. Im, K. Kuczera, T. Lazaridis, J. Ma, V. Ovchinnikov, E. Paci, R. W. Pastor, C. B. Post, J. Z. Pu, M. Schaefer, B. Tidor, R. M. Vena...

    2009

  4. [4]

    Developing a molecular dynamics force field for both folded and disordered protein states,

    P. Robustelli, S. Piana, and D. E. Shaw, “Developing a molecular dynamics force field for both folded and disordered protein states,”Proc Natl Acad Sci U S A, vol. 115, no. 21, pp. E4758–E4766, 2018

    2018

  5. [5]

    CHARMM36m: an improved force field for folded and intrinsically disordered proteins,

    J. Huang, S. Rauscher, G. Nawrocki, T. Ran, M. Feig, B. L. de Groot, H. Grubm¨ uller, and A. D. MacKerell, “CHARMM36m: an improved force field for folded and intrinsically disordered proteins,”Nature Methods, vol. 14, pp. 71–73, 2017

    2017

  6. [6]

    Building a More Predictive Protein Force Field: A Systematic and Reproducible Route to AMBER-FB15,

    L.-P. Wang, K. A. McKiernan, J. Gomes, K. A. Beauchamp, T. Head-Gordon, J. E. Rice, W. C. Swope, T. J. Mart ´ ınez, and V. S. Pande, “Building a More Predictive Protein Force Field: A Systematic and Reproducible Route to AMBER-FB15,”The Journal of Physical Chemistry B, vol. 121, no. 16, pp. 4023–4039, 2017

    2017

  7. [7]

    Development and Benchmarking of Open Force Field 2.0.0: The Sage Small Molecule Force Field,

    S. Boothroyd, P. K. Behara, O. C. Madin, D. F. Hahn, H. Jang, V. Gapsys, J. R. Wagner, J. T. Horton, D. L. Dotson, M. W. Thompson, J. Maat, T. Gokey, L.-P. Wang, D. J. Cole, M. K. Gilson, J. D. Chodera, C. I. Bayly, M. R. Shirts, and D. L. Mobley, “Development and Benchmarking of Open Force Field 2.0.0: The Sage Small Molecule Force Field,”Journal of Chem...

    2023

  8. [8]

    Systematic design of biomolecular force fields,

    D. van der Spoel, “Systematic design of biomolecular force fields,”Curr Opin Struct Biol, vol. 67, pp. 18–24, 2021

    2021

Show all 87 references
  1. [9]

    Accurate machine learning force fields via experimental and simulation data fusion,

    S. R¨ ocken and J. Zavadlav, “Accurate machine learning force fields via experimental and simulation data fusion,” npj Computational Materials, vol. 10, no. 69, 2024

    2024

  2. [10]

    Toward empirical force fields that match experimental observables,

    T. Fr¨ ohlking, M. Bernetti, N. Calonaci, and G. Bussi, “Toward empirical force fields that match experimental observables,”Journal of Chemical Physics, vol. 152, no. 23, p. 230902, 2020

    2020

  3. [11]

    On the design space between molecular mechanics and machine learning force fields,

    Y. Wang, K. Takaba, M. S. Chen, M. Wieder, Y. Xu, T. Zhu, J. Z. H. Zhang, A. Nagle, K. Yu, X. Wang, D. J. Cole, J. A. Rackers, K. Cho, J. G. Greener, P. Eastman, S. Martiniani, and M. E. Tuckerman, “On the design space between molecular mechanics and machine learning force fie...

    2025

  4. [12]

    How fast-folding proteins fold,

    K. Lindorff-Larsen, S. Piana, R. O. Dror, and D. E. Shaw, “How fast-folding proteins fold,”Science, vol. 334, no. 6055, pp. 517–520, 2011

    2011

  5. [13]

    Accurate and Reliable Prediction of Relative Ligand Binding Potency in Prospective Drug Discovery by Way of a Modern Free-Energy Calculation Protocol and Force Field,

    L. Wang, Y. Wu, Y. Deng, B. Kim, L. Pierce, G. Krilov, D. Lupyan, S. Robinson, M. K. Dahlgren, J. Greenwood, D. L. Romero, C. Masse, J. L. Knight, T. Steinbrecher, T. Beuming, W. Damm, E. Harder, W. Sherman, M. Brewer, R. Wester, M. Murcko, L. Frye, R. Farid, T. Lin, D. L. Mob...

    2015

  6. [14]

    Large- scale collaborative assessment of binding free energy calculations for drug discovery using OpenFE,

    H. M. Baumann, J. T. Horton, M. M. Henry, A. Travitz, B. Ries, R. J. Gowers, D. W. H. Swenson, I. Pulido, D. Rufa, D. L. Dotson, N. Bansal, J. P. Bluck, H. Broughton, K. Campbell, L. Cao, B. Frieg, V. Gapsys, H. G¨ oddeke, M. Kl¨ ahn, S. K. Lakkaraju, S. M. Linker, T. L¨ ohr, ...

    2025

  7. [15]

    Machine- learned molecular mechanics force fields from large-scale quantum chemical data,

    K. Takaba, A. J. Friedman, C. Cavender, P. K. Behara, I. Pulido, M. M. Henry, H. MacDermott-Opeskin, C. R. Iacovella, A. M. Nagle, A. M. Payne, M. R. Shirts, D. L. Mobley, J. D. Chodera, and Y. Wang, “Machine- learned molecular mechanics force fields from large-scale quantum c...

    2024

  8. [16]

    End-to-end differentiable construction of molecular mechanics force fields,

    Y. Wang, J. Fass, B. Kaminow, J. E. Herr, D. Rufa, I. Zhang, I. Pulido, M. Henry, H. E. Bruce Macdonald, K. Tak- aba, and J. D. Chodera, “End-to-end differentiable construction of molecular mechanics force fields,”Chemical Science, vol. 13, no. 41, pp. 12016–12033, 2022

    2022

  9. [17]

    Regularized by Physics: Graph Neural Network Parametrized Potentials for the Description of Intermolecular Interactions,

    M. Th¨ urlemann, L. B¨ oselt, and S. Riniker, “Regularized by Physics: Graph Neural Network Parametrized Potentials for the Description of Intermolecular Interactions,”Journal of Chemical Theory and Computation, vol. 19, no. 2, pp. 562–579, 2023

    2023

  10. [18]

    Grappa – a machine learned molecular mechanics force field,

    L. Seute, E. Hartmann, J. St¨ uhmer, and F. Gr¨ ater, “Grappa – a machine learned molecular mechanics force field,” Chemical Science, vol. 16, pp. 2907–2930, 2025

    2025

  11. [19]

    Con- densation of Force Field Parameters from Machine Learning Predicted Distributions for High-Throughput Virtual Screening Applications,

    D. Bonanni, Y. Zhang, D. Gadioli, G. Scarpellini, P. Morerio, A. Del Bue, A. R. Beccari, and G. Palermo, “Con- densation of Force Field Parameters from Machine Learning Predicted Distributions for High-Throughput Virtual Screening Applications,”Journal of Chemical Information ...

    2025

  12. [20]

    Advancing Force Fields Pa- rameterization: A Directed Graph Attention Networks Approach,

    G. Chen, T. Jaffrelot Inizan, T. Pl´ e, L. Lagard` ere, J.-P. Piquemal, and Y. Maday, “Advancing Force Fields Pa- rameterization: A Directed Graph Attention Networks Approach,”Journal of Chemical Theory and Computation, vol. 20, no. 13, pp. 5558–5569, 2024

    2024

  13. [21]

    Atom typing using graph representation learning: How do models learn chemistry?,

    J. Zhang, “Atom typing using graph representation learning: How do models learn chemistry?,”Journal of Chemical Physics, vol. 156, no. 20, p. 204108, 2022

    2022

  14. [22]

    Data-driven parametrization of molecular mechanics force fields for expansive chemical space coverage,

    T. Zheng, A. Wang, X. Han, Y. Xia, X. Xu, J. Zhan, Y. Liu, Y. Chen, Z. Wang, X. Wu, S. Gong, and W. Yan, “Data-driven parametrization of molecular mechanics force fields for expansive chemical space coverage,”Chemical Science, vol. 16, pp. 2730–2740, 2025

    2025

  15. [23]

    Bridging Quantum Mechanics to Organic Liquid Properties via a Universal Force Field,

    T. Zheng, X. Xu, Z. Wang, Z. Yang, Y. Wang, X. Han, Z. Mu, Z. Zhang, S. Liu, S. Gong, K. Yu, and W. Yan, “Bridging Quantum Mechanics to Organic Liquid Properties via a Universal Force Field,”arXiv, vol. 2508.08575, 2025

    2025 arXiv

  16. [24]

    Building Force Fields: An Automatic, Systematic, and Reproducible Approach,

    L.-P. Wang, T. J. Martinez, and V. S. Pande, “Building Force Fields: An Automatic, Systematic, and Reproducible Approach,”J Phys Chem Lett, vol. 5, no. 11, pp. 1885–1891, 2014

    2014

  17. [25]

    Learning neural network potentials from experimental data via Differentiable Trajectory Reweighting,

    S. Thaler and J. Zavadlav, “Learning neural network potentials from experimental data via Differentiable Trajectory Reweighting,”Nat Commun, vol. 12, no. 6884, 2021

    2021

  18. [26]

    A double exponential potential for van der Waals interaction,

    X. Wu and B. R. Brooks, “A double exponential potential for van der Waals interaction,”AIP Advances, vol. 9, p. 065304, 2019

    2019

  19. [27]

    A transferable double exponential potential for condensed phase simulations of small molecules,

    J. T. Horton, S. Boothroyd, P. K. Behara, D. L. Mobley, and D. J. Cole, “A transferable double exponential potential for condensed phase simulations of small molecules,”Digital Discovery, vol. 2, pp. 1178–1187, 2023

    2023

  20. [28]

    Accurate and Efficient Corrections for Missing Dispersion Interactions in Molecular Simulations,

    M. R. Shirts, D. L. Mobley, J. D. Chodera, and V. S. Pande, “Accurate and Efficient Corrections for Missing Dispersion Interactions in Molecular Simulations,”The Journal of Physical Chemistry B, vol. 111, no. 45, pp. 13052– 13063, 2007

    2007

  21. [29]

    On the Role of London Dispersion Forces in Biomolecular Structure Determi- nation,

    M. Kol´ aˇ r, T. Kubaˇ r, and P. Hobza, “On the Role of London Dispersion Forces in Biomolecular Structure Determi- nation,”The Journal of Physical Chemistry B, vol. 115, no. 24, pp. 8038–8046, 2011

    2011

  22. [30]

    The PHAST 2.0 Force Field for General Small Molecule and Materials Simu- lations,

    A. Hogan, L. Ritter, and B. Space, “The PHAST 2.0 Force Field for General Small Molecule and Materials Simu- lations,”Journal of Chemical Theory and Computation, vol. 21, no. 12, pp. 6034–6047, 2025

    2025

  23. [31]

    Perspective: Advances and challenges in treating van der Waals dispersion forces in density functional theory,

    J. Klimeˇ s and A. Michaelides, “Perspective: Advances and challenges in treating van der Waals dispersion forces in density functional theory,”Journal of Chemical Physics, vol. 137, p. 120901, 2012

    2012

  24. [32]

    Optimized Lennard-Jones Parameters for Druglike Small Molecules,

    E. Boulanger, L. Huang, C. Rupakheti, A. D. J. MacKerell, and B. Roux, “Optimized Lennard-Jones Parameters for Druglike Small Molecules,”Journal of Chemical Theory and Computation, vol. 14, no. 6, pp. 3121–3131, 2018. 26

    2018

  25. [33]

    Beyond Born–Mayer: Improved Models for Short-Range Repulsion in ab Initio Force Fields,

    M. J. Van Vleet, A. J. Misquitta, A. J. Stone, and J. R. Schmidt, “Beyond Born–Mayer: Improved Models for Short-Range Repulsion in ab Initio Force Fields,”Journal of Chemical Theory and Computation, vol. 12, no. 8, pp. 3851–3870, 2016

    2016

  26. [34]

    Accelerating Development and Execution Speed with Just-in-Time GPU Code Gener- ation,

    P. Eastman and V. Pande, “Accelerating Development and Execution Speed with Just-in-Time GPU Code Gener- ation,”GPU Computing Gems Jade Edition, pp. 399–407, 2012

    2012

  27. [35]

    OpenMM 7: Rapid development of high performance algorithms for molecular dynamics,

    P. Eastman, J. Swails, J. D. Chodera, R. T. McGibbon, Y. Zhao, K. A. Beauchamp, L. P. Wang, A. C. Simmonett, M. P. Harrigan, C. D. Stern, R. P. Wiewiora, B. R. Brooks, and V. S. Pande, “OpenMM 7: Rapid development of high performance algorithms for molecular dynamics,”PLoS Com...

    2017

  28. [36]

    Minimal Basis Iterative Stockholder: Atoms in Molecules for Force-Field Development,

    T. Verstraelen, S. Vandenbrande, F. Heidar-Zadeh, L. Vanduyfhuys, V. Van Speybroeck, M. Waroquier, and P. W. Ayers, “Minimal Basis Iterative Stockholder: Atoms in Molecules for Force-Field Development,”Journal of Chem- ical Theory and Computation, vol. 12, no. 8, pp. 3894–3912, 2016

    2016

  29. [37]

    SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentials,

    P. Eastman, P. K. Behara, D. L. Dotson, R. Galvelis, J. E. Herr, J. T. Horton, Y. Mao, J. D. Chodera, B. P. Pritchard, Y. Wang, G. De Fabritiis, and T. E. Markland, “SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentials,”Scientific Data,...

    2023

  30. [38]

    Nutmeg and SPICE: Models and Data for Biomolecular Machine Learning,

    P. Eastman, B. P. Pritchard, J. D. Chodera, and T. E. Markland, “Nutmeg and SPICE: Models and Data for Biomolecular Machine Learning,”Journal of Chemical Theory and Computation, vol. 20, no. 19, pp. 8583–8593, 2024

    2024

  31. [39]

    MACE-OFF: Short-Range Transferable Machine Learning Force Fields for Organic Molecules,

    D. P. Kov´ acs, J. H. Moore, N. J. Browning, I. Batatia, J. T. Horton, Y. Pu, V. Kapil, W. C. Witt, I.-B. Magd˘ au, D. J. Cole, and G. Cs´ anyi, “MACE-OFF: Short-Range Transferable Machine Learning Force Fields for Organic Molecules,”Journal of the American Chemical Society, v...

    2025

  32. [40]

    QC Optimization Dataset: OpenFF Industry Benchmark Season 1 v1.2,

    J. T. Horton, D. Dotson, J. Clark, J. Mitchell, L. Wang, and D. Cole, “QC Optimization Dataset: OpenFF Industry Benchmark Season 1 v1.2,”Zenodo, 2025

    2025

  33. [41]

    TFD: torsion fingerprints as a new measure to compare small molecule conformations,

    T. Schulz-Gasch, C. Scharfer, W. Guba, and M. Rarey, “TFD: torsion fingerprints as a new measure to compare small molecule conformations,”Journal of Chemical Information and Modeling, vol. 52, no. 6, pp. 1499–1512, 2012

    2012

  34. [42]

    Development of a Force Field for the Simulation of Single-Chain Proteins and Protein-Protein Complexes,

    S. Piana, P. Robustelli, D. Tan, S. Chen, and D. E. Shaw, “Development of a Force Field for the Simulation of Single-Chain Proteins and Protein-Protein Complexes,”Journal of Chemical Theory and Computation, vol. 16, no. 4, pp. 2494–2507, 2020

    2020

  35. [43]

    Differentiable simulation to develop molecular dynamics force fields for disordered proteins,

    J. G. Greener, “Differentiable simulation to develop molecular dynamics force fields for disordered proteins,”Chem- ical Science, vol. 15, no. 13, pp. 4897–4909, 2024

    2024

  36. [44]

    Molecular Basis of Small-Molecule Binding toα-Synuclein,

    P. Robustelli, A. Ibanez-de Opakua, C. Campbell-Bezat, F. Giordanetto, S. Becker, M. Zweckstetter, A. C. Pan, and D. E. Shaw, “Molecular Basis of Small-Molecule Binding toα-Synuclein,”Journal of the American Chemical Society, vol. 144, no. 6, pp. 2501–2510, 2022

    2022

  37. [45]

    Comparison of simple potential functions for simulating liquid water,

    W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey, and M. L. Klein, “Comparison of simple potential functions for simulating liquid water,”Journal of Chemical Physics, vol. 79, pp. 926–935, 1983

    1983

  38. [46]

    A Graph Neural Network Charge Model Targeting Accurate Electrostatic Properties of Organic Molecules,

    C. Adams, J. T. Horton, L. Wang, S. Boothroyd, D. L. Mobley, D. W. Wright, and D. J. Cole, “A Graph Neural Network Charge Model Targeting Accurate Electrostatic Properties of Organic Molecules,”Journal of Chemical Theory and Computation, vol. 21, no. 23, pp. 12133–12148, 2025

    2025

  39. [47]

    Open Force Field Evaluator: An Au- tomated, Efficient, and Scalable Framework for the Estimation of Physical Properties from Molecular Simulation,

    S. Boothroyd, L.-P. Wang, D. L. Mobley, J. D. Chodera, and M. R. Shirts, “Open Force Field Evaluator: An Au- tomated, Efficient, and Scalable Framework for the Estimation of Physical Properties from Molecular Simulation,” Journal of Chemical Theory and Computation, vol. 18, no...

    2022

  40. [48]

    The representation of van der Waals (vdW) interactions in molecular mechanics force fields: potential form, combination rules, and vdW parameters,

    T. A. Halgren, “The representation of van der Waals (vdW) interactions in molecular mechanics force fields: potential form, combination rules, and vdW parameters,”Journal of the American Chemical Society, vol. 114, no. 20, pp. 7827–7843, 1992

    1992

  41. [49]

    The maximal and current accuracy of rigorous protein-ligand binding free energy calculations,

    G. A. Ross, C. Lu, G. Scarabelli, S. K. Albanese, E. Houang, R. Abel, E. D. Harder, and L. Wang, “The maximal and current accuracy of rigorous protein-ligand binding free energy calculations,”Communications Chemistry, vol. 6, no. 1, p. 222, 2023

    2023

  42. [50]

    Best Practices for Constructing, Preparing, and Evaluating Protein- Ligand Binding Affinity Benchmarks [Article v1.0],

    D. Hahn, C. Bayly, M. L. Boby, H. Bruce Macdonald, J. Chodera, V. Gapsys, A. Mey, D. Mobley, L. Perez Benito, C. Schindler, G. Tresadern, and G. Warren, “Best Practices for Constructing, Preparing, and Evaluating Protein- Ligand Binding Affinity Benchmarks [Article v1.0],”Livi...

    2022

  43. [51]

    Bang for your Buck: Trading Off Utility with Time and Money for Non-Equilibrium Switching in Orion: https://www.youtube.com/watch?v=d76MZoxc2eo,

    C. Bayly, “Bang for your Buck: Trading Off Utility with Time and Money for Non-Equilibrium Switching in Orion: https://www.youtube.com/watch?v=d76MZoxc2eo,” 2024. 27

    2024

  44. [52]

    Influence of the Lennard-Jones Combination Rules on the Simulated Prop- erties of Organic Liquids at Optimal Force-Field Parametrization,

    M. P. Oliveira and P. H. H¨ unenberger, “Influence of the Lennard-Jones Combination Rules on the Simulated Prop- erties of Organic Liquids at Optimal Force-Field Parametrization,”Journal of Chemical Theory and Computation, vol. 19, no. 7, pp. 2048–2063, 2023

  45. [53]

    New combining rules for rare gas van der waals parameters,

    M. Waldman and A. Hagler, “New combining rules for rare gas van der waals parameters,”Journal of Computational Chemistry, vol. 14, no. 9, pp. 1077–1084, 1993

    1993

  46. [54]

    Inductive Representation Learning on Large Graphs,

    W. L. Hamilton, R. Ying, and J. Leskovec, “Inductive Representation Learning on Large Graphs,”31st Conference on Neural Information Processing Systems (NIPS 2017), 2017

    2017

  47. [55]

    Open Force Field BespokeFit: Automat- ing Bespoke Torsion Parametrization at Scale,

    J. T. Horton, S. Boothroyd, J. Wagner, J. A. Mitchell, T. Gokey, D. L. Dotson, P. K. Behara, V. K. Ramaswamy, M. Mackey, J. D. Chodera, J. Anwar, D. L. Mobley, and D. J. Cole, “Open Force Field BespokeFit: Automat- ing Bespoke Torsion Parametrization at Scale,”Journal of Chemi...

    2022

  48. [56]

    The Open Force Field Initiative: Open Software and Open Science for Molecular Modeling,

    L. Wang, P. K. Behara, M. W. Thompson, T. Gokey, Y. Wang, J. R. Wagner, D. J. Cole, M. K. Gilson, M. R. Shirts, and D. L. Mobley, “The Open Force Field Initiative: Open Software and Open Science for Molecular Modeling,” The Journal of Physical Chemistry B, vol. 128, no. 29, pp...

    2024

  49. [57]

    Improved Treatment of 1–4 Interactions in Force Fields for Molecular Dynamics Simulations,

    A. S. Abdullah, Y. Wang, M. F. Menger, S. Sami, and T. Head-Gordon, “Improved Treatment of 1–4 Interactions in Force Fields for Molecular Dynamics Simulations,”Journal of Chemical Theory and Computation, vol. 21, no. 22, pp. 11669–11678, 2025

    2025

  50. [58]

    Force field development phase II: Relaxation of physics-based criteria... or inclusion of more rigorous physics into the representation of molecular energetics,

    A. T. Hagler, “Force field development phase II: Relaxation of physics-based criteria... or inclusion of more rigorous physics into the representation of molecular energetics,”Journal of Computer-Aided Molecular Design, vol. 33, pp. 205–264, 2019

    2019

  51. [59]

    Tuning Potential Functions to Host-Guest Binding Data,

    J. Setiadi, S. Boothroyd, D. R. Slochower, D. L. Dotson, M. W. Thompson, J. R. Wagner, L.-P. Wang, and M. K. Gilson, “Tuning Potential Functions to Host-Guest Binding Data,”Journal of Chemical Theory and Computation, vol. 20, no. 1, pp. 239–252, 2024

    2024

  52. [60]

    Fine-tuning molecular mechanics force fields to experimental free energy measurements,

    D. Rufa, J. Fass, and J. D. Chodera, “Fine-tuning molecular mechanics force fields to experimental free energy measurements,”bioRxiv, 2025

    2025

  53. [61]

    Statistically optimal analysis of samples from multiple equilibrium states,

    M. R. Shirts and J. D. Chodera, “Statistically optimal analysis of samples from multiple equilibrium states,”Journal of Chemical Physics, vol. 129, p. 124105, 2008

    2008

  54. [62]

    Analyzing Atomic Interactions in Molecules as Learned by Neural Networks,

    M. Esders, T. Schnake, J. Lederer, A. Kabylda, G. Montavon, A. Tkatchenko, and K.-R. M¨ uller, “Analyzing Atomic Interactions in Molecules as Learned by Neural Networks,”Journal of Chemical Theory and Computation, vol. 21, no. 2, pp. 714–729, 2025

    2025

  55. [63]

    Escaping Atom Types in Force Fields Using Direct Chemical Perception,

    D. L. Mobley, C. C. Bannan, A. Rizzi, C. I. Bayly, J. D. Chodera, V. T. Lim, N. M. Lim, K. A. Beauchamp, D. R. Slochower, M. R. Shirts, M. K. Gilson, and P. K. Eastman, “Escaping Atom Types in Force Fields Using Direct Chemical Perception,”Journal of Chemical Theory and Comput...

    2018

  56. [64]

    Janossy Pooling: Learning Deep Permutation-Invariant Functions for Variable-Size Inputs,

    R. L. Murphy, B. Srinivasan, V. Rao, and B. Ribeiro, “Janossy Pooling: Learning Deep Permutation-Invariant Functions for Variable-Size Inputs,”International Conference on Learning Representations, 2019

    2019

  57. [65]

    Robustness in the fitting of molecular mechanics parameters,

    K. Vanommeslaeghe, M. Yang, and A. D. MacKerell, “Robustness in the fitting of molecular mechanics parameters,” Journal of Computational Chemistry, vol. 36, pp. 1083–1101, 2015

    2015

  58. [66]

    Biomolecular dynamics with machine-learned quantum- mechanical force fields trained on diverse chemical fragments,

    O. T. Unke, M. St¨ ohr, S. Ganscha, T. Unterthiner, H. Maennel, S. Kashubin, D. Ahlin, M. Gastegger, L. M. San- donas, J. T. Berryman, A. Tkatchenko, and K.-R. M¨ uller, “Biomolecular dynamics with machine-learned quantum- mechanical force fields trained on diverse chemical fr...

    2024

  59. [67]

    Machine learning force fields for molecular liquids: Ethylene Carbonate/Ethyl Methyl Carbonate binary solvent,

    I.-B. Magd˘ au, D. J. Arismendi-Arrieta, H. E. Smith, C. P. Grey, K. Hermansson, and G. Cs´ anyi, “Machine learning force fields for molecular liquids: Ethylene Carbonate/Ethyl Methyl Carbonate binary solvent,”npj Computational Materials, vol. 9, p. 146, 2023

    2023

  60. [68]

    Structure-Based Experimental Datasets for Benchmarking Protein Simulation Force Fields [Article v1.0],

    C. E. Cavender, D. A. Case, J. C.-H. Chen, L. T. Chong, D. A. Keedy, K. Lindorff-Larsen, D. L. Mobley, O. H. S. Ollila, C. Oostenbrink, P. J. Robustelli, and et al., “Structure-Based Experimental Datasets for Benchmarking Protein Simulation Force Fields [Article v1.0],”Living ...

    2025

  61. [69]

    Enhanced Sampling Methods for Molecular Dynamics Simulations [Article v1.0],

    J. H´ enin, T. Leli` evre, M. R. Shirts, O. Valsson, and L. Delemotte, “Enhanced Sampling Methods for Molecular Dynamics Simulations [Article v1.0],”Living Journal of Computational Molecular Science, vol. 4, no. 1, p. 1583, 2022

    2022

  62. [70]

    Biological Magnetic Resonance Data Bank,

    J. C. Hoch, K. Baskaran, H. Burr, J. Chin, H. R. Eghbalnia, T. Fujiwara, M. R. Gryk, T. Iwata, C. Kojima, G. Kurisu, D. Maziuk, Y. Miyanoiri, J. R. Wedell, C. Wilburn, H. Yao, and M. Yokochi, “Biological Magnetic Resonance Data Bank,”Nucleic Acids Research, vol. 51, no. D1, pp...

    2023

  63. [71]

    Predicting chemical shifts with graph neural networks,

    Z. Yang, M. Chakraborty, and A. D. White, “Predicting chemical shifts with graph neural networks,”Chemical Science, vol. 12, pp. 10802–10809, 2021

    2021

  64. [72]

    Julia: A fresh approach to numerical computing,

    J. Bezanson, A. Edelman, S. Karpinski, and V. B. Shah, “Julia: A fresh approach to numerical computing,”SIAM Review, vol. 59, no. 1, pp. 65–98, 2017

    2017

  65. [73]

    Julia for biologists,

    E. Roesch, J. G. Greener, A. L. MacLean, H. Nassar, C. Rackauckas, T. E. Holy, and M. P. H. Stumpf, “Julia for biologists,”Nat Methods, vol. 20, no. 5, pp. 655–664, 2023

    2023

  66. [74]

    Instead of Rewriting Foreign Code for Machine Learning, Automatically Synthesize Fast Gradients,

    W. S. Moses and V. Churavy, “Instead of Rewriting Foreign Code for Machine Learning, Automatically Synthesize Fast Gradients,”Advances in Neural Information Processing Systems, vol. 33, pp. 12472–12485, 2020

    2020

  67. [75]

    Don’t Unroll Adjoint: Differentiating SSA-Form Programs,

    M. Innes, “Don’t Unroll Adjoint: Differentiating SSA-Form Programs,”arXiv, vol. 1810.07951, 2018

    2018 arXiv

  68. [76]

    Particle mesh Ewald: An N·log(N) method for Ewald sums in large systems,

    T. Darden, D. York, and L. Pedersen, “Particle mesh Ewald: An N·log(N) method for Ewald sums in large systems,” Journal of Chemical Physics, vol. 98, pp. 10089–10092, 1993

    1993

  69. [77]

    MDAnalysis: A Python Package for the Rapid Analysis of Molecular Dynamics Simulations,

    R. J. Gowers, M. Linke, J. Barnoud, T. J. E. Reddy, M. N. Melo, S. L. Seyler, J. Doma´ nski, D. L. Dotson, S. Buchoux, I. M. Kenney, and O. Beckstein, “MDAnalysis: A Python Package for the Rapid Analysis of Molecular Dynamics Simulations,”Proceedings of the 15th Python in Scie...

    2016

  70. [78]

    BioStructures.jl: read, write and manipulate macromolecular structures in Julia,

    J. G. Greener, J. Selvaraj, and B. J. Ward, “BioStructures.jl: read, write and manipulate macromolecular structures in Julia,”Bioinformatics, vol. 36, no. 14, pp. 4206–4207, 2020

    2020

  71. [79]

    PyTorch: An Imperative Style, High-Performance Deep Learning Library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Kopf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala, “PyTorch: An Imperative Style, High-Pe...

    2019

  72. [80]

    RDKit: Open-source cheminformatics. https://www.rdkit.org,

    “RDKit: Open-source cheminformatics. https://www.rdkit.org,” 2025

    2025

  73. [81]

    Long-time-step molecular dynamics through hydrogen mass repartitioning,

    C. W. Hopkins, S. Le Grand, R. C. Walker, and A. E. Roitberg, “Long-time-step molecular dynamics through hydrogen mass repartitioning,”Journal of Chemical Theory and Computation, vol. 11, no. 4, pp. 1864–1874, 2015

    2015

  74. [82]

    proteinbenchmark: https://github.com/openforcefield/proteinbenchmark,

    C. E. Cavender, “proteinbenchmark: https://github.com/openforcefield/proteinbenchmark,” 2022

    2022

  75. [83]

    OpenFreeEnergy/IndustryBenchmarks2024: v1.0.0,

    H. Baumann, I. Alibay, J. Horton, B. Ries, M. Henry, W. C. Swope, Lili-Cao, L. Tsidilkovski, steinbt2, V. Gapsys, nupurbansal, l045340, benefrg, alexanderhwilliams, T. L¨ ohr, M. Kl¨ ahn, MCompChem, J. Bluck, C. Stern, and A. Travitz, “OpenFreeEnergy/IndustryBenchmarks2024: v1...

    2025

  76. [84]

    Lead optimization mapper: automating free energy calculations for lead optimization,

    S. Liu, Y. Wu, T. Lin, R. Abel, J. P. Redmann, C. M. Summa, V. R. Jaber, N. M. Lim, and D. L. Mobley, “Lead optimization mapper: automating free energy calculations for lead optimization,”Journal of Computer-Aided Molecular Design, vol. 27, no. 9, pp. 755–770, 2013

    2013

  77. [85]

    Kartograf: A Geometrically Accurate Atom Mapper for Hybrid-Topology Relative Free Energy Calculations,

    B. Ries, I. Alibay, D. W. H. Swenson, H. M. Baumann, M. M. Henry, J. R. B. Eastwood, and R. J. Gowers, “Kartograf: A Geometrically Accurate Atom Mapper for Hybrid-Topology Relative Free Energy Calculations,” Journal of Chemical Theory and Computation, vol. 20, no. 5, pp. 1862–...

    2024

  78. [86]

    Optimal Measurement Network of Pairwise Differences,

    H. Xu, “Optimal Measurement Network of Pairwise Differences,”Journal of Chemical Information and Modeling, vol. 59, no. 11, pp. 4720–4728, 2019

    2019

  79. [87]

    Fast, efficient generation of high-quality atomic charges. AM1-BCC model: I. Method,

    A. Jakalian, B. L. Bush, D. B. Jack, and C. I. Bayly, “Fast, efficient generation of high-quality atomic charges. AM1-BCC model: I. Method,”Journal of Computational Chemistry, vol. 21, no. 2, pp. 132–146, 2000. 29 Training a force field for proteins and small molecules from sc...

    2000

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.