Using structure to evaluate the accuracy and transferability of machine learning potentials of biomolecules
- 1Department of Biomolecular and Chemical Engineering, North Carolina State University Raleigh North Carolina 27607 USA
- 2Department of Chemistry, North Carolina State University Raleigh North Carolina 27607 USA
Abstract
Molecular simulation is a powerful tool to describe chemical and physical processes across length scales. Fundamental tradeoffs between the accuracy of quantum mechanics and speed of classical models makes simulating biosystems very difficult. Machine learning potentials (MLPs) offer a potential solution which combines the accuracy of electronic structure theory with the computational scaling of classical approaches. Applying MLPs to biosystems could unlock new developments in drug discovery, biological mechanisms and interactions. Despite this promise, rigorous benchmarks are needed to determine whether MLPs reliably reproduce first-principles counterparts. It is commonplace to use energies and forces as the primary performance metrics for MLPs. While essential, we show that models with among the lowest energy errors can simultaneously produce the largest structural errors. Through simulations of two isomeric biomolecules, this work shows that structural information must be added as a metric to assess both the overall accuracy and transferability of an MLP. Here, we show that when structural quantities are included alongside traditional metrics, MLPs can achieve high fidelity to first-principles reference data and remain transferable across related chemical motifs examined. These findings underscore the importance of structure-based validation and support the continued use of MLPs for biomolecular simulations.
Related
- Concentration-Dependent Speciation in ZnCl2 Electrolyte Solutions: Transferability of Machine-Learning Interatomic Potentials and Speciation Sensitivity of EXAFS
- Fine-tuning MLIP foundation models: strategies for accuracy and transferability
- Guidelines to Evaluate Structure-based Methods for Protein-Ligand Binding Prediction
- Accuracy and Efficiency Benchmarks of Pretrained Machine Learning Potentials for Molecular Simulations
- Triplet Pruning and Spatial Decomposition in M3GNet: Accuracy and Memory Trade-offs in Scaling Graph Neural Network Potentials
BibTeX
@article{biberic2026using,
title = {{Using structure to evaluate the accuracy and transferability of machine learning potentials of biomolecules}},
author = {Lejla Š. Biberić and Nisarg Joshi and Jim Pfaendtner},
journal = {RSC Advances},
year = {2026},
month = oct,
doi = {10.1039/d6ra05346g},
url = {https://doi.org/10.1039/d6ra05346g},
note = {CC BY-NC 3.0},
}
RIS
TY - JOUR
TI - Using structure to evaluate the accuracy and transferability of machine learning potentials of biomolecules
AU - Lejla Š. Biberić
AU - Nisarg Joshi
AU - Jim Pfaendtner
PY - 2026
DA - 2026/10/08
JO - RSC Advances
DO - 10.1039/d6ra05346g
UR - https://doi.org/10.1039/d6ra05346g
AB - Molecular simulation is a powerful tool to describe chemical and physical processes across length scales. Fundamental tradeoffs between the accuracy of quantum mechanics and speed of classical models makes simulating biosystems very difficult. Machine learning potentials (MLPs) offer a potential solution which combines the accuracy of electronic structure theory with the computational scaling of classical approaches. Applying MLPs to biosystems could unlock new developments in drug discovery, biological mechanisms and interactions. Despite this promise, rigorous benchmarks are needed to determine whether MLPs reliably reproduce first-principles counterparts. It is commonplace to use energies and forces as the primary performance metrics for MLPs. While essential, we show that models with among the lowest energy errors can simultaneously produce the largest structural errors. Through simulations of two isomeric biomolecules, this work shows that structural information must be added as a metric to assess both the overall accuracy and transferability of an MLP. Here, we show that when structural quantities are included alongside traditional metrics, MLPs can achieve high fidelity to first-principles reference data and remain transferable across related chemical motifs examined. These findings underscore the importance of structure-based validation and support the continued use of MLPs for biomolecular simulations.
SN - 2046-2069
ER -