Exploring the World of Lipids with Machine-Learned Coarse-Grained Models
Abstract
Lipid bilayers are central to cellular organization and function, yet their atomistic modelling remains constrained by its high computational cost. While coarse-grained (CG) models can overcome this bottleneck, they are limited in accuracy due to the use of simplified functional forms based on molecular mechanics. Using a hybrid functional form combining physics-based terms with a graph neural network, we present implicit-solvent, machine-learned CG models for POPC. When trained on single-lipid data, the model recovers the atomistic conformational landscape. By adding short atomistic simulations of a small flat bilayer to the training data and combining force matching with a denoising objective, we train a model that accurately reproduces bilayer structure and dynamics, in closer agreement with the atomistic reference than existing models, like Martini3 and DryMartini. Beyond its training data, the model stabilizes curved geometries, including vesicles, and captures complex cooperative phenomena such as bilayer and vesicle self-assembly and pore sealing. While limitations remain, particularly regarding chemical transferability, our results establish machine-learned coarse-grained potentials as a promising path towards robust and accurate CG models of lipids and membranes.
Related
- Hessian Matching for Machine-Learned Coarse-Grained Molecular Dynamics
- Phase separating microdomains in highly coarse-grained lipids
- ESPResSo++: A Fast and Extensible Molecular Simulation Package for Coarse-Grained Models
- Coarse-grained models for loop quantum gravity and their renormalization
- A Gaussian process coarse-grained potential for Na-montmorillonite
BibTeX
@misc{pasostrejo2026exploring,
title = {{Exploring the World of Lipids with Machine-Learned Coarse-Grained Models}},
author = {Aldo S. Pasos-Trejo and Daniel P. Ramirez-Echemendia and Daniel Conde-Torres and D. Peter Tieleman and Cecilia Clementi},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.05100},
eprint = {2611.05100},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.05100},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Exploring the World of Lipids with Machine-Learned Coarse-Grained Models
AU - Aldo S. Pasos-Trejo
AU - Daniel P. Ramirez-Echemendia
AU - Daniel Conde-Torres
AU - D. Peter Tieleman
AU - Cecilia Clementi
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2611.05100
UR - https://arxiv.org/abs/2611.05100
AB - Lipid bilayers are central to cellular organization and function, yet their atomistic modelling remains constrained by its high computational cost. While coarse-grained (CG) models can overcome this bottleneck, they are limited in accuracy due to the use of simplified functional forms based on molecular mechanics. Using a hybrid functional form combining physics-based terms with a graph neural network, we present implicit-solvent, machine-learned CG models for POPC. When trained on single-lipid data, the model recovers the atomistic conformational landscape. By adding short atomistic simulations of a small flat bilayer to the training data and combining force matching with a denoising objective, we train a model that accurately reproduces bilayer structure and dynamics, in closer agreement with the atomistic reference than existing models, like Martini3 and DryMartini. Beyond its training data, the model stabilizes curved geometries, including vesicles, and captures complex cooperative phenomena such as bilayer and vesicle self-assembly and pore sealing. While limitations remain, particularly regarding chemical transferability, our results establish machine-learned coarse-grained potentials as a promising path towards robust and accurate CG models of lipids and membranes.
ER -