Skip to content
DraconiticResearch
Search Sources API
190,332 studies31 sourcesUpdated 23 min ago
PreprintCC BYChemRxiv

Atomic Environment Aware Lennard-Jones Parameterization for Electrostatic Embedding ML/MM Simulations

  • João Morado1,2
  • Kirill Zinovjev3,4
  • Lester O. Hedges5
  • Julien Michel1,2
  • Daniel J. Cole6,7
7 affiliations
  1. 1EaStCHEM School of Chemistry
  2. 2University of Edinburgh
  3. 3Departament de Química Física
  4. 4Universitat de València
  5. 5OpenBioSim Community Interest Company
  6. 6School of Natural and Environmental Sciences
  7. 7Newcastle University
Open at sourcePDFDOICite

Abstract

Hybrid machine learning/molecular mechanics (ML/MM) methods have emerged as a powerful approach to multiscale modelling. Recent advances have focused on electrostatic embedding schemes, in which atoms in the ML region are polarized by the surrounding MM environment. Consequently, ML/MM electrostatics have been the subject of extensive study, whereas dispersion and close-range repulsive interactions remain comparatively underexplored. In analogy to established QM/MM practice, exchange-repulsion and dispersion interactions are commonly described using Lennard-Jones (LJ) 12-6 potentials with parameters transferred from existing force field libraries. However, such transferable parameters may be suboptimal for ML/MM interfaces, where consistency with the electrostatic embedding model is desirable. Here, we develop methodologies for deriving dynamic LJ parameters based on the exchange-hole dipole moment (XDM) model and the Fedorov polarizability scaling relation. Moreover, we propose a one-shot procedure for fine-tuning LJ parameters to experimental free energies. Focusing on the electrostatic machine learning embedding (EMLE) scheme, we show that these approaches yield EMLE models with competitive accuracy relative to conventional force fields and mechanical embedding schemes, while additionally providing chemically environment-aware LJ parameters. This capability offers a route to improved accuracy in demanding applications such as reactive modelling, where dispersion coefficients may depend strongly on the local chemical environment.

Related

  1. A Hybrid Message-Passing and Electric-Field Representation for Electrostatic QM/MM Embedding with Machine Learning Interatomic Potentials
    Raphael T. Husistein, Shobhit S. Chaturvedi, Thomas Weymuth +5
    arXivPreprintarXiv licenceFull textPhysical chemistryBiological & medical physicsarXiv
  2. Machine-Learned Short-Range Repulsive Hamiltonians for Electrostatic QM/MM Embedding
    Minghao Cao, Jinfeng Liu
    ChemRxivPreprintCC BYFull textAtomic, molecular & optical physicsChemRxiv
  3. Influence of Electrostatic Environment on the Proximity Spin-Orbit Coupling in Graphene on Transition-Metal Dichalcogenides
    Bert Jorissen, Bart Partoens, Aires Ferreira +1
    arXivPreprintarXiv licenceFull textCondensed matterarXiv
  4. Efficient kinetic Monte Carlo simulations with long-range electrostatic interactions
    Roya Ebrahimi Viand, Bat-Amgalan Bat-Erdene, Karsten Reuter +1
    arXivPreprintarXiv licenceFull textStatistical & nonlinear physicsarXiv
  5. EasyHybrid: An Interactive Graphical Environment for Quantum, Classical and Hybrid Simulations with pDynamo3
    José Fernando Ruggiero Bachega, Gustavo Hagen, Carlos Sequeiros-Borja +4
    Journal of Chemical Information and ModelingPeer-reviewedCC BY 4.0Full textAtomic, molecular & optical physicsHAL

Cite

BibTeXRIS
BibTeX
@misc{morado2026atomic,
  title = {{Atomic Environment Aware Lennard-Jones Parameterization for Electrostatic Embedding ML/MM Simulations}},
  author = {João Morado and Kirill Zinovjev and Lester O. Hedges and Julien Michel and Daniel J. Cole},
  howpublished = {ChemRxiv},
  year = {2026},
  month = sep,
  doi = {10.26434/chemrxiv.15004404/v2},
  url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15004404/v2},
  note = {CC BY},
}
RIS
TY  - UNPB
TI  - Atomic Environment Aware Lennard-Jones Parameterization for Electrostatic Embedding ML/MM Simulations
AU  - João Morado
AU  - Kirill Zinovjev
AU  - Lester O. Hedges
AU  - Julien Michel
AU  - Daniel J. Cole
PY  - 2026
DA  - 2026/09/22
PB  - ChemRxiv
DO  - 10.26434/chemrxiv.15004404/v2
UR  - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15004404/v2
AB  - Hybrid machine learning/molecular mechanics (ML/MM) methods have emerged as a powerful approach to multiscale modelling. Recent advances have focused on electrostatic embedding schemes, in which atoms in the ML region are polarized by the surrounding MM environment. Consequently, ML/MM electrostatics have been the subject of extensive study, whereas dispersion and close-range repulsive interactions remain comparatively underexplored. In analogy to established QM/MM practice, exchange-repulsion and dispersion interactions are commonly described using Lennard-Jones (LJ) 12-6 potentials with parameters transferred from existing force field libraries. However, such transferable parameters may be suboptimal for ML/MM interfaces, where consistency with the electrostatic embedding model is desirable. Here, we develop methodologies for deriving dynamic LJ parameters based on the exchange-hole dipole moment (XDM) model and the Fedorov polarizability scaling relation. Moreover, we propose a one-shot procedure for fine-tuning LJ parameters to experimental free energies. Focusing on the electrostatic machine learning embedding (EMLE) scheme, we show that these approaches yield EMLE models with competitive accuracy relative to conventional force fields and mechanical embedding schemes, while additionally providing chemically environment-aware LJ parameters. This capability offers a route to improved accuracy in demanding applications such as reactive modelling, where dispersion coefficients may depend strongly on the local chemical environment.
ER  -