Pith. sign in

REVIEW 4 major objections 3 minor 44 references

QuantumBind-RBFE: Accurate Relative Binding Free Energy Calculations Using Neural Network Potentials

T0 review · 4 major / 3 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a released neural network potential, AceFF 1.0, improves relative binding free energy predictions over GAFF2 and reaches ranking correlations comparable to OPLS4 when used as the ligand potential in an NNP/MM scheme.

desk verdict Careful, reproducible benchmark of a new NNP for RBFE; the improvement over GAFF2 is plausible but the reported bootstrap intervals overlap and the abstract overstates it. read the letter →

arxiv 2501.01811 v2 pith:TQ4OMVFU submitted 2025-01-03 physics.chem-ph cs.LGphysics.comp-ph

Francesc Sabanés Zariquiey , Stephen E. Farr , Stefan Doerr , Gianni De Fabritiis This is my paper · ORCID
classification physics.chem-phcs.LGphysics.comp-ph
keywords relativebindingfreeenergyneuralnetworkpotentialsAceFFTensorNetalchemicaltransfermethodNNP/MMforcefieldaccuracydrugdiscovery
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 tests whether a released neural network potential, AceFF 1.0, can serve as the ligand force field in relative binding free energy (RBFE) calculations. Using the Alchemical Transfer Method with an NNP/MM mechanical-embedding setup, the authors report that AceFF improves accuracy and ranking over GAFF2 on the seven-target JACS benchmark (RMSE 0.99 vs 1.17 kcal/mol, MAE 0.79 vs 0.90, Kendall tau 0.59 vs 0.55). Against OPLS4 with FEP+, AceFF shows slightly larger errors but comparable correlations (Kendall tau 0.59 vs 0.66), and the authors note the comparison mixes different simulation engines and protocols. The same simulations run stably at 2 fs, twice the usual NNP timestep, with accuracy close to 1 fs runs. If these results hold, a publicly available neural potential can be dropped into drug-discovery workflows without the torsion-scan parameterization of classical force fields.

What carries the argument

The carrying object is AceFF 1.0, a one-layer TensorNet neural network potential: an equivariant message-passing architecture that represents atomic environments with Cartesian tensor features and predicts the ligand's intramolecular energy and forces. It is embedded through the NNP/MM energy expression $V = V_{\text{NNP}}(\mathbf{r}_{\text{NNP}}) + V_{\text{MM}}(\mathbf{r}_{\text{MM}}) + V_{\text{NNP-MM}}(\mathbf{r})$, in which the coupling term is the classical electrostatic and van der Waals interactions between the ligand and the surrounding protein and solvent. The free energy difference is computed with the Alchemical Transfer Method, which alchemically transfers the ligand between bound and unbound states in a dual-topology simulation box. The NNP replaces all bonded ligand terms, including torsions, so no torsion-scan fitting is needed; the paper notes that range-corrected variants that include short-range environment interactions exist but are not used here.

What would settle it

Replace the classical coupling term in the same NNP/MM scheme with a polarizable or range-corrected quantum treatment and rerun the seven-target benchmark; if RMSE falls well below 0.99 kcal/mol, the remaining error is dominated by the mechanical-embedding coupling rather than by the ligand's intramolecular potential, weakening the paper's attribution of the improvement to AceFF. A cheaper check is to benchmark a target rich in halogen bonds or salt bridges, where fixed-charge electrostatics are known to struggle.

Watch

Extended reading notes

Core claim

The central claim is that AceFF 1.0, a TensorNet-based neural network potential trained on quantum-chemical energies and forces, is accurate enough to replace classical ligand force fields in RBFE calculations. In the QuantumBind-RBFE protocol, the total energy is split as $V = V_{\text{NNP}}(\mathbf{r}_{\text{NNP}}) + V_{\text{MM}}(\mathbf{r}_{\text{MM}}) + V_{\text{NNP-MM}}(\mathbf{r})$, where only the ligand's intramolecular energy is computed by the NNP and all ligand-environment interactions are classical. Across seven protein targets and 280 perturbation edges, the model outperforms GAFF2 run under the identical protocol, reduces outliers, and improves compound prioritization on most targets. The paper attributes the gain to better treatment of the ligand's internal energetics, especially strain, since the sampled conformational distributions are similar to GAFF2. AceFF 1.0 is stable at a 2 fs timestep, and the paper demonstrates that 2 fs and 1 fs runs give nearly identical errors and correlations.

Load-bearing premise

The load-bearing premise is that all ligand–protein and ligand–water interactions are accurately described by fixed-charge classical molecular mechanics, with the neural network correcting only the ligand's internal energy; if polarization, charge transfer, or other quantum effects across the binding interface matter, the NNP cannot fix those errors.

Editorial extensions

If this is right

  • Publicly available neural potentials can replace classical ligand force fields in RBFE calculations, eliminating the need for torsion-scan parameterization.
  • AceFF's stable 2 fs timestep doubles simulation speed relative to earlier NNP models, making NNP/MM free-energy campaigns practical on existing hardware.
  • Ranking of the most active compounds, not just error metrics, improves over GAFF2 on most of the tested targets, which is the metric that matters in hit-to-lead and lead optimization.
  • Ligands outside the NNP's training distribution, such as those with a -2 charge, fail badly (RMSE above 3 kcal/mol and negative Kendall tau on PTP1B), so chemical coverage of charge states is a limiting factor.

Reading between the lines

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

  • Because the OPLS4 comparison uses a different molecular dynamics engine and free-energy protocol, the small accuracy gap (RMSE 0.99 vs 0.78 kcal/mol) is not a clean force-field comparison; an identical-protocol study could narrow or reverse it.
  • The paper switched from AM1BCC to RESP charges compared with its earlier ANI-2x work, so part of the improvement over that baseline may come from the charge model rather than the neural network itself; an ablation with identical charges would separate the two effects.
  • If the 2 fs stability generalizes to larger, flexible ligands, NNP/MM costs approach classical MM at 4 fs, which would make neural-potential FEP a viable screening tool rather than a retrospective benchmark.
  • The PTP1B failure is a natural stress test: retraining the model on -2 and +2 charged species would provide a direct test of whether the charge-range limitation, rather than the mechanical-embedding coupling, is the main barrier to broader applicability.
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. The manuscript presents QuantumBind-RBFE, an alchemical relative binding free energy protocol that uses the AceFF 1.0 neural network potential for the ligand within an NNP/MM mechanical-embedding scheme and the Alchemical Transfer Method (ATM). AceFF 1.0 is benchmarked on seven targets of the JACS dataset (280 edges; PTP1B is explicitly excluded because its dianionic ligands lie outside the trained charge range, though a stress test on PTP1B is also reported). The paper claims improved accuracy and correlation over GAFF2 under the same protocol (RMSE 0.99 vs 1.17 kcal/mol, MAE 0.79 vs 0.90, Kendall tau 0.59 vs 0.55), comparable ranking but slightly worse RMSE/MAE than OPLS4/FEP+ (a cross-protocol comparison), better accuracy than ANI-2x, and stable 2 fs timestep operation. The authors conclude that AceFF 1.0 can already serve as a practical ligand force field for RBFE calculations while noting limitations for rare charge states and the mechanical-embedding approximation.

Significance. If the headline claims are accepted, this is a practically important result: it would demonstrate that a publicly released neural network potential can replace a classical ligand force field in an established alchemical RBFE workflow, with meaningful speed advantages over earlier NNP/MM implementations. The study has genuine strengths: triplicate 70 ns/replica ATM simulations, UWHAM analysis, 95% bootstrap confidence intervals, a same-protocol GAFF2 control, and public availability of the model, input structures, and analysis scripts. The additional PTP1B stress test, run outside the model's trained charge range, is a useful falsifiability check. However, the central quantitative claim of improved accuracy over GAFF2 is not supported by the reported statistics alone, because the bootstrap confidence intervals overlap for every headline metric and no paired significance test is provided. The mechanical-embedding assumption in Eq. (1) is also acknowledged but not stress-tested, which limits the generality of the conclusions.

major comments (4)
  1. [§3.1, Table S2] The claim that AceFF 1.0 shows 'improved accuracy' over GAFF2 is not statistically established. For the all-target ΔG metrics, the 95% bootstrap intervals overlap for RMSE ([0.89,1.10] vs [1.03,1.30]), MAE ([0.71,0.89] vs [0.79,1.01]), and Kendall tau ([0.52,0.65] vs [0.48,0.62]). The same is true for the ΔΔG metrics in Table S4 (RMSE [1.10,1.33] vs [1.30,1.62], MAE [0.85,1.03] vs [0.97,1.20], tau [0.40,0.51] vs [0.34,0.49]). Since the abstract and conclusions assert 'overall improved accuracy and correlation', the authors should either report a paired significance test (e.g., a paired bootstrap over ligands or edges, or a permutation test on per-edge ΔΔG errors) or soften the claim to 'comparable to, and in some targets better than, GAFF2'. Given the data and code are public, such a test is straightforward and would resolve whether the point-estimate differences are sampling noise or a genuine force-field effect.
  2. [§2.5] The bootstrap resampling unit is not specified. The reported confidence intervals are used to support the main comparisons, but the ΔG metrics are computed over ligands and the ΔΔG metrics over edges, and edges share ligands through the perturbation network. If the bootstrap resamples individual edges rather than independent units (ligands, or clusters of correlated edges), the intervals will be too narrow and the overlap assessment in the previous comment would be optimistic. Please state the resampling unit explicitly and, if edges are resampled, justify why dependencies are negligible or switch to a cluster bootstrap by ligand.
  3. [§2.1, Eq. (1)] The mechanical-embedding assumption that all ligand-protein and ligand-solvent nonbonded interactions are adequately described by fixed-charge classical MM (RESP charges, TIP3P, ff14SB) is the load-bearing modeling choice, but it is not stress-tested. The paper attributes AceFF's improvements over GAFF2 mainly to better treatment of ligand internal energetics, which is consistent with this choice, yet the claim that AceFF is a 'viable alternative for RBFE calculations' in general goes beyond what can be concluded from the JACS systems, where polarization and charge-transfer effects may be small for the tested ligands. I recommend either adding a limited test (e.g., comparison against a range-corrected or explicitly polarizable scheme on a subset) or explicitly restricting the conclusion to systems where the fixed-charge environment approximation is adequate.
  4. [§3.3, Tables S9-S10] The claim that '2 fs timestep simulations demonstrate comparable accuracy to 1 fs runs' rests on point estimates whose differences are not tested for significance. For example, Table S9 shows TYK2 ΔG RMSE of 0.47 at 1 fs versus 0.74 at 2 fs, with overlapping intervals ([0.23,0.66] vs [0.38,0.81]); similar differences appear for THROMBIN (0.57 vs 0.80). The stability at 2 fs is convincing, but 'comparable accuracy' would be better supported by reporting paired per-target differences or by presenting the claim as 'within the observed bootstrap uncertainty' rather than as equivalence.
minor comments (3)
  1. [§2.2 and Supporting Information] The manuscript states that PTP1B is omitted from the evaluation, but Tables S3, S5, S6, and S7 report PTP1B rows for AceFF 1.0 and the charge-model comparison. The role of these rows should be clarified; if they are part of the 'going beyond the limits' stress test, they should be labeled as such in the table captions to avoid confusion with the main benchmark.
  2. [§2.4] The sentence 'We used the Amber ff14SB parameters as well as the TIP3P water model' appears without a citation in the body text; the ff14SB reference is present in the reference list but should be cited at the point of use, as done for TIP3P elsewhere.
  3. [§3.1] The discussion of the MCL1 decrease in Kendall tau (0.28 for AceFF vs 0.44 for GAFF2) says it is 'primarily due to the overprediction of binding affinities for five ligands', but no quantitative support for this attribution is provided; a sentence identifying those five ligands and their predicted versus experimental values would make the claim verifiable.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: AceFF is trained on DFT data and benchmarked against external experimental binding affinities; self-citations are methodological, not load-bearing reductions.

full rationale

The paper's central claim is an empirical benchmark result, not a derivation. AceFF 1.0 is trained on Acellera's proprietary DFT energies and forces (wB97M-V/def2-tzvppd), and no parameter is fitted to the experimental binding free energies used for evaluation. The RBFE predictions are genuine out-of-sample tests against the JACS dataset and CCSD(T)/CBS torsion references. Equation 1 defines the NNP/MM mechanical-embedding scheme, but this is a stated modeling choice rather than a hidden definition of the target quantity; the coupling term is computed with fixed-charge MM, and the paper honestly lists this as a limitation. The self-citations to AceFF, NNP/MM, ATM, and prior ANI-2x work describe tools and earlier validations, but the load-bearing comparison with GAFF2 is recomputed in this work under the same protocol, and the OPLS4 comparison uses independently published FEP+ results. The acknowledged PTP1B failure is an explicit out-of-distribution test, not a circular retrodiction. Bootstrap confidence intervals overlapping between AceFF and GAFF2 is a statistical-significance concern, not a circularity concern: the reported difference may be noise, but nothing about the construction forces the improvement. No fitted input is renamed as a prediction, no uniqueness theorem is imported from the authors, and no known result is merely relabeled. Therefore the derivation chain is self-contained against external benchmarks, and the circularity score is 0.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the fitted AceFF model and the fixed ATM protocol, but no binding data was used to fit either. The load-bearing axioms are the adequacy of mechanical embedding, the accuracy of the JACS experimental references, the DFT level used for training, and the statistical estimators. No new physical entities are introduced.

free parameters (2)
  • AceFF 1.0 neural network weights = not disclosed (proprietary training set)
    The entire RBFE accuracy claim rests on this NNP, trained on Acellera's internal dataset at wB97M-V/def2-tzvppd; the paper does not describe the architecture size, training set composition, or hyperparameters beyond 'TensorNet 1-layer'.
  • Softplus alchemical potential parameters (alpha, u0, w0) = alpha=0.1 kcal/mol^-1, u0=110 kcal/mol, w0=0
    Hand-chosen ATM protocol constants from Table S1; not fitted to the benchmark, but they define the alchemical pathway used in all runs and could affect convergence.
assumptions (4)
  • domain assumption Mechanical embedding (Eq. 1) is a sufficient model: ligand-environment interactions are accurately described by MM with fixed charges.
    Invoked in Section 2.1; if polarization or charge transfer matters, the NNP's internal accuracy cannot recover the error.
  • domain assumption Experimental binding affinities in the JACS dataset are accurate reference values.
    Used as ground truth for all RMSE, MAE, and Kendall metrics; any error in these values propagates to all methods equally.
  • domain assumption wB97M-V/def2-tzvppd is an adequate reference level of theory for training AceFF.
    AceFF is trained on forces and energies at this level (Section 2.3); if the level is inadequate for ligand strain, the RBFE errors follow.
  • standard math UWHAM and cinnabar maximum-likelihood estimators are unbiased for these alchemical samples.
    Used in Section 2.5 for free energy estimation; assumed valid under the sampling protocol.

how reviews work

0 comments
Cite this review

Pith. "Pith review of QuantumBind-RBFE: Accurate Relative Binding Free Energy Calculations Using Neural Network Potentials." pith.science (2026). https://pith.science/paper/TQ4OMVFU

@misc{pith2026250101811,
  author       = {Pith},
  title        = {Pith review of: QuantumBind-RBFE: Accurate Relative Binding Free Energy Calculations Using Neural Network Potentials},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TQ4OMVFU}},
  note         = {Machine review of arXiv:2501.01811}
}
read the original abstract

Accurate prediction of protein-ligand binding affinities is crucial in drug discovery, particularly during hit-to-lead and lead optimization phases, however, limitations in ligand force fields continue to impact prediction accuracy. In this work, we validate relative binding free energy (RBFE) accuracy using neural network potentials (NNPs) for the ligands. We utilize a novel NNP model, AceFF 1.0, based on the TensorNet architecture for small molecules that broadens the applicability to diverse drug-like compounds, including all important chemical elements and supporting charged molecules. Using established benchmarks, we show overall improved accuracy and correlation in binding affinity predictions compared with GAFF2 for molecular mechanics and ANI2-x for NNPs. Slightly less accuracy but comparable correlations with OPLS4. We also show that we can run the NNP simulations at 2 fs timestep, at least two times larger than previous NNP models, providing significant speed gains. The results show promise for further evolutions of free energy calculations using NNPs while demonstrating its practical use already with the current generation. The code and NNP model are publicly available for research use.

Figures

Figures reproduced from arXiv: 2501.01811 by the authors.

Figure 1
Figure 1. Description of the NNP/MM scheme. While the ligand is simulated with a neural net￾work potential (NNP), the rest of the system is treated with classical molecular mechanics (MM) We utilized the Alchemical Transfer Method (ATM) to perform the RBFE calculations. 29 This methodology has been previously vali￾dated across multiple benchmarks, 30,31 demon￾strating its reliability for accurate binding free energy predictio… view at source ↗
Figure 2
Figure 2. The AceFF model is among the best performing and comparable to MACE but at a much faster computational speed. AceFF is us￾ing a faster TensorNet 1-layer model, instead of the 2-layer model generally used in17 for maxi￾mum accuracy, which has a speed of 80ns/day for a small molecule19 (timestep 1fs) compared to MACE-OFF23-Small with 7.5ns/day 13 . 2.4 QuantumBind-RBFE Calcu￾lation details The workflow in this project… view at source ↗
Figure 3
Figure 3. (Left) Root Mean Squared Error (RMSE) and (right) Kendall tau correlation for the [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Scatterplots of predicted ∆G values for each evaluated system using AceFF 1.0. The green and yellow shaded areas represent absolute error thresholds of 1 kcal/mol and 2 kcal/mol, respectively. Additional metrics, including mean absolute error (MAE), root mean square er…
Figure 5
Figure 5. Figure 5: Identification of top compounds across datasets. The left plot shows the accuracy in [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Comparison of AceFF 1.0 model ac￾curacy at 1fs and 2fs timesteps. The top panel shows Root Mean Squared Error (RMSE) and the bottom panel shows Kendall tau correla￾tion for ∆G values across a subset of systems. While runs at both timesteps yield similar per￾formance, t…
Figure 8
Figure 8. Figure 8: Left: Example on one of the ligands of the PTP1B dataset, highlighting the -1 charge at [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 37 canonical work pages

  1. [1]

    Jorgensen, W. L. The many roles of computation in drug discovery. Science 2004, 303, 1813--1818

    work page 2004

  2. [2]

    D.; Berne, B.; Friesner, R

    Abel, R.; Wang, L.; Harder, E. D.; Berne, B.; Friesner, R. A. Advancing drug discovery through enhanced free energy calculations. Acc. Chem. Res. 2017, 50, 1625--1632

    work page 2017

  3. [3]

    A.; Riniker, S.; Cournia, Z

    Armacost, K. A.; Riniker, S.; Cournia, Z. Novel Directions in Free Energy Methods and Applications. J. Chem. Inf. Model. 2020, 60, 1--5, PMID: 31983210

    work page 2020

  4. [4]

    York, D. M. Modern alchemical free energy methods for drug discovery explained. ACS Physical Chemistry Au 2023, 3, 478--491

    work page 2023

  5. [5]

    Alchemical Transformations and Beyond: Recent Advances and Real-World Applications of Free Energy Calculations in Drug Discovery

    Qian, R.; Xue, J.; Xu, Y.; Huang, J. Alchemical Transformations and Beyond: Recent Advances and Real-World Applications of Free Energy Calculations in Drug Discovery. Journal of Chemical Information and Modeling 2024, 64, 7214--7237

    work page 2024

  6. [6]

    M.; Caldwell, J

    Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development and testing of a general amber force field. Journal of computational chemistry 2004, 25, 1157--1174

    work page 2004

  7. [7]

    A.; Case, D

    Wang, J.; Wang, W.; Kollman, P. A.; Case, D. A. Automatic atom type and bond type perception in molecular mechanical calculations. J. Mol. Graphics Modell. 2006, 25, 247--260

    work page 2006

  8. [8]

    Journal of computational chemistry 2010, 31, 671--690

    Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, A., Igor MacKerell Jr CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields. Journal of computational chemistry 2010, 31, 671--690

    work page 2010

Show all 44 references
  1. [9]

    P.; MacKerell Jr, A

    Vanommeslaeghe, K.; Raman, E. P.; MacKerell Jr, A. D. Automation of the CHARMM General Force Field (CGenFF) II: assignment of bonded parameters and partial atomic charges. Journal of chemical information and modeling 2012, 52, 3155--3168

    2012

  2. [10]

    G.; Boothroyd, S.; Jang, H.; Hahn, D

    Qiu, Y.; Smith, D. G.; Boothroyd, S.; Jang, H.; Hahn, D. F.; Wagner, J.; Bannan, C. C.; Gokey, T.; Lim, V. T.; Stern, C. D.; others Development and benchmarking of open force field v1. 0.0—the parsley small-molecule force field. Journal of chemical theory and computation 2021,...

    2021

  3. [11]

    K.; Madin, O

    Boothroyd, S.; Behara, P. K.; Madin, O. C.; Hahn, D. F.; Jang, H.; Gapsys, V.; Wagner, J. R.; Horton, J. T.; Dotson, D. L.; Thompson, M. W.; others Development and benchmarking of open force field 2.0. 0: the Sage small molecule force field. Journal of chemical theory and comp...

    2023

  4. [12]

    V.; Joshi, C

    Duval, A.; Mathis, S. V.; Joshi, C. K.; Schmidt, V.; Miret, S.; Malliaros, F. D.; Cohen, T.; Lio, P.; Bengio, Y.; Bronstein, M. A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systems. arXiv preprint arXiv:2312.07511 2023,

    2023 arXiv

  5. [13]

    P.; Moore, J

    Kov \'a cs, D. P.; Moore, J. H.; Browning, N. J.; Batatia, I.; Horton, J. T.; Kapil, V.; Witt, W. C.; Magd a u, I.-B.; Cole, D. J.; Cs \'a nyi, G. MACE-OFF23: Transferable machine learning force fields for organic molecules. arXiv preprint arXiv:2312.15211 2023,

    2023 arXiv

  6. [14]

    S.; Huddleston, K

    Devereux, C.; Smith, J. S.; Huddleston, K. K.; Barros, K.; Zubatyuk, R.; Isayev, O.; Roitberg, A. E. Extending the Applicability of the ANI Deep Learning Molecular Potential to Sulfur and Halogens. Journal of Chemical Theory and Computation 2020, 16, 4192--4202, PMID: 32543858

    2020

  7. [15]

    AIMNet2: a neural network potential to meet your neutral, charged, organic, and elemental-organic needs

    Anstine, D.; Zubatyuk, R.; Isayev, O. AIMNet2: a neural network potential to meet your neutral, charged, organic, and elemental-organic needs. ChemrXiv preprint chemrxiv-2023-296ch-v3 2024,

    2023

  8. [16]

    Torchmd-net: equivariant transformers for neural network based molecular potentials

    Th \"o lke, P.; De Fabritiis, G. Torchmd-net: equivariant transformers for neural network based molecular potentials. arXiv preprint arXiv:2202.02541 2022,

    2022 arXiv

  9. [17]

    TensorNet: Cartesian Tensor Representations for Efficient Learning of Molecular Potentials

    Simeon, G.; De Fabritiis, G. TensorNet: Cartesian Tensor Representations for Efficient Learning of Molecular Potentials. Advances in Neural Information Processing Systems. 2023; pp 37334--37353

    2023

  10. [18]

    TorchMD: A deep learning framework for molecular simulations

    Doerr, S.; Majewski, M.; P \'e rez, A.; Kramer, A.; Clementi, C.; Noe, F.; Giorgino, T.; De Fabritiis, G. TorchMD: A deep learning framework for molecular simulations. Journal of chemical theory and computation 2021, 17, 2355--2363

    2021

  11. [19]

    P.; Simeon, G.; Galvelis, R.; Mirarchi, A.; Eastman, P.; Doerr, S.; Th \"o lke, P.; Markland, T

    Pelaez, R. P.; Simeon, G.; Galvelis, R.; Mirarchi, A.; Eastman, P.; Doerr, S.; Th \"o lke, P.; Markland, T. E.; De Fabritiis, G. TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulations. Journal of Chemical Theory and Computation 2024,

    2024

  12. [20]

    https://huggingface.co/Acellera/AceForce-1.0, 2024; Accessed: 2024-12-30

    Acellera AceForce-1.0 on Hugging Face. https://huggingface.co/Acellera/AceForce-1.0, 2024; Accessed: 2024-12-30

    2024

  13. [21]

    Machine Learning Potentials: A Roadmap Toward Next-Generation Biomolecular Simulations

    De Fabritiis, G. Machine Learning Potentials: A Roadmap Toward Next-Generation Biomolecular Simulations. arXiv preprint arXiv:2408.12625 2024,

    2024 arXiv

  14. [22]

    D.; Markland, T

    Sabane\'es Zariquiey, F.; Galvelis, R.; Gallicchio, E.; Chodera, J. D.; Markland, T. E.; De Fabritiis, G. Enhancing Protein--Ligand Binding Affinity Predictions Using Neural Network Potentials. Journal of Chemical Information and Modeling 2024, 64, 1481--1485

    2024

  15. [23]

    E.; Chodera, J

    Galvelis, R.; Varela-Rial, A.; Doerr, S.; Fino, R.; Eastman, P.; Markland, T. E.; Chodera, J. D.; De Fabritiis, G. NNP/MM: Accelerating molecular dynamics simulations with machine learning potentials and molecular mechanics. Journal of chemical information and modeling 2023, 6...

    2023

  16. [24]

    Z.; Pal, R

    Azimi, S.; Khuttan, S.; Wu, J. Z.; Pal, R. K.; Gallicchio, E. Relative binding free energy calculations for ligands with diverse scaffolds with the alchemical transfer method. J. Chem. Inf. Model. 2022, 62, 309--323

    2022

  17. [25]

    Validation of the Alchemical Transfer Method for the Estimation of Relative Binding Affinities of Molecular Series

    Saban\'es Zariquiey, F.; P\'esrez, A.; Majewski, M.; Gallicchio, E.; De Fabritiis, G. Validation of the Alchemical Transfer Method for the Estimation of Relative Binding Affinities of Molecular Series. Journal of Chemical Information and Modeling 2023, 63, 2438--2444, PMID: 37042797

    2023

  18. [26]

    Performance and Analysis of the Alchemical Transfer Method for Binding Free Energy Predictions of Diverse Ligands

    Chen, L.; Wu, Y.; Wu, C.; Silveira, A.; Sherman, W.; Xu, H.; Gallicchio, E. Performance and Analysis of the Alchemical Transfer Method for Binding Free Energy Predictions of Diverse Ligands. 2023

    2023

  19. [27]

    K.; Greenwood, J.; Romero, D

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

    2015

  20. [28]

    A.; Dahlgren, M

    Lu, C.; Wu, C.; Ghoreishi, D.; Chen, W.; Wang, L.; Damm, W.; Ross, G. A.; Dahlgren, M. K.; Russell, E.; Von Bargen, C. D.; others OPLS4: Improving force field accuracy on challenging regimes of chemical space. Journal of chemical theory and computation 2021, 17, 4291--4300

    2021

  21. [29]

    https://www.acellera.com/, 2024; Accessed: 2024-12-30

    Acellera Acellera Homepage. https://www.acellera.com/, 2024; Accessed: 2024-12-30

    2024

  22. [30]

    D.; James, N

    Sellers, B. D.; James, N. C.; Gobbi, A. A comparison of quantum and molecular mechanical methods to estimate strain energy in druglike fragments. Journal of chemical information and modeling 2017, 57, 1265--1275

    2017

  23. [31]

    S.; Roitberg, A

    Gao, X.; Ramezanghorbani, F.; Isayev, O.; Smith, J. S.; Roitberg, A. E. TorchANI: a free and open source PyTorch-based deep learning implementation of the ANI neural network potentials. Journal of chemical information and modeling 2020, 60, 3408--3415

    2020

  24. [32]

    GFN2-xTB—An accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions

    Bannwarth, C.; Ehlert, S.; Grimme, S. GFN2-xTB—An accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions. Journal of chemical theory and computation 2019, 15, 1652--1671

    2019

  25. [33]

    Geometry optimization made simple with translation and rotation coordinates

    Wang, L.-P.; Song, C. Geometry optimization made simple with translation and rotation coordinates. The Journal of chemical physics 2016, 144

    2016

  26. [34]

    A.; Wang, L.; Wagner, J.; trevorgokey openforcefield/openff-recharge: 0.5.0

    SimonBoothroyd; Thompson, M.; Mitchell, J. A.; Wang, L.; Wagner, J.; trevorgokey openforcefield/openff-recharge: 0.5.0. 2023; https://doi.org/10.5281/zenodo.8118623

    2023 doi

  27. [35]

    M.; Harvey, M

    Galvelis, R.; Doerr, S.; Damas, J. M.; Harvey, M. J.; De Fabritiis, G. A scalable molecular force field parameterization method based on density functional theory and quantum-level machine learning. Journal of chemical information and modeling 2019, 59, 3485--3493

    2019

  28. [36]

    HTMD: high-throughput molecular dynamics for molecular discovery

    Doerr, S.; Harvey, M.; No \'e , F.; De Fabritiis, G. HTMD: high-throughput molecular dynamics for molecular discovery. Journal of chemical theory and computation 2016, 12, 1845--1852

    2016

  29. [37]

    F.; Gapsys, V.; de Groot, B

    Hahn, D. F.; Gapsys, V.; de Groot, B. L.; Mobley, D. L.; Tresadern, G. Current State of Open Source Force Fields in Protein--Ligand Binding Affinity Predictions. Journal of Chemical Information and Modeling 2024, 64, 5063--5076

    2024

  30. [38]

    Blinded prediction of protein--ligand binding affinity using Amber thermodynamic integration for the 2018 D3R grand challenge 4

    Zou, J.; Tian, C.; Simmerling, C. Blinded prediction of protein--ligand binding affinity using Amber thermodynamic integration for the 2018 D3R grand challenge 4. J. Comput.-Aided Mol. Des. 2019, 33, 1021--1029

    2018

  31. [39]

    A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; Hauser, K

    Maier, J. A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; Hauser, K. E.; Simmerling, C. ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput. 2015, 11, 3696--3713

    2015

  32. [40]

    https://www.gpugrid.net/, Accessed: 2024-12-10

    GPUGRID - Volunteer Computing for Molecular Simulations. https://www.gpugrid.net/, Accessed: 2024-12-10

    2024

  33. [41]

    P.; Abreu, C

    Eastman, P.; Galvelis, R.; Pel \'a ez, R. P.; Abreu, C. R.; Farr, S. E.; Gallicchio, E.; Gorenko, A.; Henry, M. M.; Hu, F.; Huang, J.; others OpenMM 8: molecular dynamics simulation with machine learning potentials. The Journal of Physical Chemistry B 2023, 128, 109--116

    2023

  34. [42]

    Tan, Z.; Gallicchio, E.; Lapelosa, M.; Levy, R. M. Theory of binless multi-state free energy estimation with applications to protein-ligand binding. The Journal of Chemical Physics 2012, 136, 04B608

    2012

  35. [43]

    B.; dfhahn; Henry, M.; Chodera, J.; Dotson, D.; Glass, W.; Pulido, I

    Macdonald, H. B.; dfhahn; Henry, M.; Chodera, J.; Dotson, D.; Glass, W.; Pulido, I. openforcefield/openff-arsenic: v0.2.1. 2022; https://doi.org/10.5281/zenodo.6210305

    2022 doi

  36. [44]

    JACS dataset

    Gapsys, V.; P \'e rez-Benito, L.; Aldeghi, M.; Seeliger, D.; van Vlijmen, H.; Gary, T.; de Groot, B. Large scale relative protein ligand binding affinities using non-equilibrium alchemy. 2020 mcitethebibliography main.tex0000664000000000000000000011745514735763545011264 0ustar...

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.