A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systems
Reviewed by Pithpith:2W5NKT6Bopen to challenge →
read the original abstract
Recent advances in computational modelling of atomic systems, spanning molecules, proteins, and materials, represent them as geometric graphs with atoms embedded as nodes in 3D Euclidean space. In these graphs, the geometric attributes transform according to the inherent physical symmetries of 3D atomic systems, including rotations and translations in Euclidean space, as well as node permutations. In recent years, Geometric Graph Neural Networks have emerged as the preferred machine learning architecture powering applications ranging from protein structure prediction to molecular simulations and material generation. Their specificity lies in the inductive biases they leverage - such as physical symmetries and chemical properties - to learn informative representations of these geometric graphs. In this opinionated paper, we provide a comprehensive and self-contained overview of the field of Geometric GNNs for 3D atomic systems. We cover fundamental background material and introduce a pedagogical taxonomy of Geometric GNN architectures: (1) invariant networks, (2) equivariant networks in Cartesian basis, (3) equivariant networks in spherical basis, and (4) unconstrained networks. Additionally, we outline key datasets and application areas and suggest future research directions. The objective of this work is to present a structured perspective on the field, making it accessible to newcomers and aiding practitioners in gaining an intuition for its mathematical abstractions.
This paper has not been read by Pith yet.
Forward citations
Cited by 11 Pith papers
-
Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models
OMat24 releases a new open dataset of 110M+ DFT calculations and EquiformerV2 models achieving SOTA on Matbench Discovery with F1>0.9 for stability and 20 meV/atom accuracy for formation energies.
-
XANE(3): An E(3)-Equivariant Graph Neural Network for Accurate Prediction of XANES Spectra from Atomic Structures
XANE(3) uses a custom E(3)-equivariant GNN with absorber attention, derivative matching loss, and Gaussian spectral readout to predict XANES spectra from atomic structures at 10^{-3} MSE on iron oxide data.
-
Navigating committor landscape of biomolecules with a general pairwise interaction model
A novel neural architecture based on Pairformer is introduced for learning committor functions to better capture dynamical features in biomolecular rare events without specialized priors.
-
Accessing gluon GTMD $F^g_{1,4}$ via the $\langle\sin(2\phi)\rangle$ azimuthal asymmetry of exclusive $\pi^0$ production in $ep$ collisions
A light-front spectator model yields the first calculation of Im(F^g_{1,4}) and the resulting sin(2φ) asymmetry in ep → epπ⁰ at EIC kinematics.
-
Adaptive Canonicalization with Application to Invariant Anisotropic Geometric Networks
Adaptive canonicalization selects input canonical forms by maximizing network predictive confidence to yield continuous symmetry-preserving models with universal approximation for equivariant geometric networks.
-
Open Materials 2024 (OMat24) Inorganic Materials Dataset and Models
OMat24 provides over 110 million DFT calculations and EquiformerV2 models that reach state-of-the-art performance on material stability and formation energy prediction.
-
Atompack: A Storage and Distribution Layer for Read-Heavy Atomistic ML Training Datasets
Atompack delivers 96x faster shuffled reads and 79% smaller artifacts than ASE LMDB baselines for complete-record atomistic ML training workloads.
-
Synthetic pre-training of graph-network models for predicting solid-state NMR parameters
Synthetic pre-training on ML-generated tensor data followed by fine-tuning on ground-truth calculations improves data efficiency for graph models of solid-state NMR parameters when the pre-training and fine-tuning dom...
-
TriForces: Augmenting Atomistic GNNs for Transferable Representations
TriForces adds a model-agnostic three-stream architecture plus self-supervised objectives to atomistic GNNs, improving transfer performance on MatBench, QM9, and limited-data OMat24 without DFT labels.
-
Enabling Structure-Only Initialization and Out-of-Distribution Generalization in GNN-based Molecular Dynamics Simulators
GNN-based MD simulators achieve stable structure-only initialization and reliable OOD generalization through inference-time physics optimization and a GNN barostat on elastic network compression tasks.
-
Atomistic Modeling of Chemical Disorder in Materials: Bridging Classical Methods and AI-Assisted Approaches
A review of classical and AI-assisted methods for modeling chemical disorder in atomistic simulations of alloys and complex materials.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.