MD-22

The MD-22 dataset is a benchmark that contains molecular dynamics simulations used to evaluate the performance of Graph Neural Networks in predicting force fields for atomistic systems.

Papers using MD-22 (14)

GFFMERGE: Efficient Merging of Graph Neural Force Fields and Beyond2026Potential Score Matching: Debiasing Molecular Structure Sampling with Potential Energy Guidance2025 · 1 citesATOM: A Pretrained Neural Operator for Multitask Molecular Dynamics2025DEQuify your force field: More efficient simulations using deep equilibrium models2025A Scalable and Quantum-Accurate Foundation Model for Biomolecular Force Field via Linearly Tensorized Quadrangle Attention2025GeoRecon: Graph-Level Representation Learning for 3D Molecules via Reconstruction-Based Pretraining2025From Molecules to Materials: Pre-training Large Generalizable Models for Atomic Property Prediction2023 · 22 citesAccurate global machine learning force fields for molecules with hundreds of atoms2022 · 9 citesLong-Short-Range Message-Passing: A Physics-Informed Framework to Capture Non-Local Interaction for Scalable Molecular Dynamics Simulation2023 · 6 citesViSNet: an equivariant geometry-enhanced graph neural network with vector-scalar interactive message passing for molecules2022 · 3 citesAtomistic Descriptor Optimization Using Complementary Euclidean and Geodesic Distance Information2024 · 3 citesNeural P$^3$M: A Long-Range Interaction Modeling Enhancer for Geometric GNNs2024 · 1 citesSE3Set: Harnessing equivariant hypergraph neural networks for molecular representation learning2024FreeCG: Free the Design Space of Clebsch-Gordan Transform for Machine Learning Force Fields2024