Abstract:Molecular dynamics (MD) simulations allow atomistic insights into chemical and biological processes. Accurate MD simulations require computationally demanding quantum-mechanical calculations, being practically limited to short timescales and few atoms. For larger systems, efficient, but much less reliable empirical force fields are used. Recently, machine learned force fields (MLFFs) emerged as an alternative means to execute MD simulations, offering similar accuracy as ab initio methods at orders-of-magnitude speedup. Until now, MLFFs mainly capture short-range interactions in small molecules or periodic materials, due to the increased complexity of constructing models and obtaining reliable reference data for large molecules, where long-ranged many-body effects become important. This work proposes a general approach to constructing accurate MLFFs for large-scale molecular simulations (GEMS) by training on "bottom-up" and "top-down" molecular fragments of varying size, from which the relevant physicochemical interactions can be learned. GEMS is applied to study the dynamics of alanine-based peptides and the 46-residue protein crambin in aqueous solution, allowing nanosecond-scale MD simulations of >25k atoms at essentially ab initio quality. Our findings suggest that structural motifs in peptides and proteins are more flexible than previously thought, indicating that simulations at ab initio accuracy might be necessary to understand dynamic biomolecular processes such as protein (mis)folding, drug-protein binding, or allosteric regulation.

Crash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Molecular Dynamics in the TEA Challenge 2023

Crash Testing Machine Learning Force Fields for Molecules, Materials, and Interfaces: Model Analysis in the TEA Challenge 2023

A foundation model for atomistic materials chemistry

Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations

Improving machine learning force fields for molecular dynamics simulations with fine-grained force metrics

Molecular Simulations with a Pretrained Neural Network and Universal Pairwise Force Fields

BIGDML: Towards Exact Machine Learning Force Fields for Materials

CHIPS-FF: Evaluating Universal Machine Learning Force Fields for Material Properties

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

Accurate Machine Learned Quantum-Mechanical Force Fields for Biomolecular Simulations

BIGDML—Towards accurate quantum machine learning force fields for materials

Efficient Machine Learning Force Field for Large-Scale Molecular Simulations of Organic Systems

Accurate global machine learning force fields for molecules with hundreds of atoms

MACE-OFF23: Transferable Machine Learning Force Fields for Organic Molecules

Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields

Challenges for Machine Learning Force Fields in Reproducing Potential Energy Surfaces of Flexible Molecules

Accurate interatomic force field for molecular dynamics simulation by hybridizing classical and machine learning potentials

Machine learning for molecular simulation

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

Force Field Analysis Software and Tools (FFAST): Assessing Machine Learning Force Fields Under the Microscope

Molecular Force Fields with Gradient-Domain Machine Learning (GDML): Comparison and Synergies with Classical Force Fields