A unified coarse-grained neural network potential for the dynamics of biomolecular complexes
Abstract
The behavior of biomolecular complexes is governed by chemically specific interactions and often involves rare events that atomistic molecular dynamics struggle to capture within accessible timescales. Here we introduce AMARO v2.0, a coarse-grained neural network potential that retains all heavy atoms and treats solvent implicitly. Hydrogen-aware bead types and total-charge conditioning provide a shared representation for proteins, drug-like molecules, and their complexes, including ternary protein-protein-ligand systems. The model preserves folded proteins and captures key features of protein conformational free-energy landscapes. Across 262 complexes spanning 11 targets, the learned potential retains most ligand poses, although quaternary structure reorientation is observed for the HIF2A ternary complex. In targeted case studies, AMARO v2.0 distinguishes the two protonation states of trypsin-benzamidine and reaches the crystal binding pose in 3.3-fold more trajectories than all-atom molecular mechanics at a fixed estimated GPU-time budget. The same framework also recovers a reference binding pose in TYK2 and supports binding free-energy estimation for both trypsin-benzamidine and TYK2-EJM31, although the resulting values are more energetically favorable than the experimental references. SHP2 and CDK8 systems expose remaining limitations in binding-pose recovery and site specificity, respectively. Together, these results support AMARO v2.0 as a shared heavy-atom potential for biomolecular complex dynamics.
Related
- From coarse-grained molecular dynamics simulations to continuum models of biomolecular condensation
- Towards transferable explicit-solvent coarse-grained models for biomolecular condensates
- AlphaBridge: Tools for the analysis of predicted biomolecular complexes
- A unified intermediate-resolution framework for modeling biomolecular condensates at near-cellular complexity
- Disulfide Crosslink Density is a Key Determinant in the Formation, Dynamics, and Maturation of Cysteine-rich Designer Biomolecular Condensates.
BibTeX
@misc{mirarchi2026unified,
title = {{A unified coarse-grained neural network potential for the dynamics of biomolecular complexes}},
author = {Antonio Mirarchi and Stephen E. Farr and Adrià Peréz and Gianni De Fabritiis},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.37606},
eprint = {2610.37606},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.37606},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - A unified coarse-grained neural network potential for the dynamics of biomolecular complexes
AU - Antonio Mirarchi
AU - Stephen E. Farr
AU - Adrià Peréz
AU - Gianni De Fabritiis
PY - 2026
DA - 2026/10/27
PB - arXiv
DO - 10.48550/arxiv.2610.37606
UR - https://arxiv.org/abs/2610.37606
AB - The behavior of biomolecular complexes is governed by chemically specific interactions and often involves rare events that atomistic molecular dynamics struggle to capture within accessible timescales. Here we introduce AMARO v2.0, a coarse-grained neural network potential that retains all heavy atoms and treats solvent implicitly. Hydrogen-aware bead types and total-charge conditioning provide a shared representation for proteins, drug-like molecules, and their complexes, including ternary protein-protein-ligand systems. The model preserves folded proteins and captures key features of protein conformational free-energy landscapes. Across 262 complexes spanning 11 targets, the learned potential retains most ligand poses, although quaternary structure reorientation is observed for the HIF2A ternary complex. In targeted case studies, AMARO v2.0 distinguishes the two protonation states of trypsin-benzamidine and reaches the crystal binding pose in 3.3-fold more trajectories than all-atom molecular mechanics at a fixed estimated GPU-time budget. The same framework also recovers a reference binding pose in TYK2 and supports binding free-energy estimation for both trypsin-benzamidine and TYK2-EJM31, although the resulting values are more energetically favorable than the experimental references. SHP2 and CDK8 systems expose remaining limitations in binding-pose recovery and site specificity, respectively. Together, these results support AMARO v2.0 as a shared heavy-atom potential for biomolecular complex dynamics.
ER -