Abstract:Abstract Effective molecular representation learning is of great importance to facilitate molecular property prediction, which is a fundamental task for the drug and material industry. Recent advances in graph neural networks (GNNs) have shown great promise in applying GNNs for molecular representation learning. Moreover, a few recent studies have also demonstrated successful applications of self-supervised learning methods to pre-train the GNNs to overcome the problem of insufficient labeled molecules. However, existing GNNs and pre-training strategies usually treat molecules as topological graph data without fully utilizing the molecular geometry information. Whereas, the three-dimensional (3D) spatial structure of a molecule, a.k.a molecular geometry, is one of the most critical factors for determining molecular physical, chemical, and biological properties. To this end, we propose a novel Geometry Enhanced Molecular representation learning method (GEM) for Chemical Representation Learning (ChemRL). At first, we design a geometry-based GNN architecture that simultaneously models atoms, bonds, and bond angles in a molecule. To be specific, we devised double graphs for a molecule: The first one encodes the atom-bond relations; The second one encodes bond-angle relations. Moreover, on top of the devised GNN architecture, we propose several novel geometry-level self-supervised learning strategies to learn spatial knowledge by utilizing the local and global molecular 3D structures. We compare ChemRL-GEM with various state-of-the-art (SOTA) baselines on different molecular benchmarks and exhibit that ChemRL-GEM can significantly outperform all baselines in both regression and classification tasks. For example, the experimental results show an overall improvement of 8.8% on average compared to SOTA baselines on the regression tasks, demonstrating the superiority of the proposed method.

Geometry-enhanced Pre-training on Interatomic Potentials

Geometry-enhanced pretraining on interatomic potentials

Robust training of machine learning interatomic potentials with dimensionality reduction and stratified sampling

Learning Interatomic Potentials at Multiple Scales

Learning together: Towards foundation models for machine learning interatomic potentials with meta-learning

Learning Together: Towards foundational models for machine learning interatomic potentials with meta-learning

Accelerating the Training and Improving the Reliability of Machine-Learned Interatomic Potentials for Strongly Anharmonic Materials through Active Learning

Geometry-aware Line Graph Transformer Pre-training for Molecular Property Prediction

Geometry-enhanced molecular representation learning for property prediction

Data-efficient multi-fidelity training for high-fidelity machine learning interatomic potentials

Improving Molecular Dynamics Simulations for Solid‐Liquid Interface with Machine Learning Interatomic Potentials

Pushing the Limits of All-Atom Geometric Graph Neural Networks: Pre-Training, Scaling and Zero-Shot Transfer

Molecular geometric deep learning

Machine learning interatomic potential: Bridge the gap between small-scale models and realistic device-scale simulations

3D Molecular Pretraining via Localized Geometric Generation

3D Interaction Geometric Pre-training for Molecular Relational Learning

ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Prediction

Online Test-time Adaptation for Interatomic Potentials

Supervised Pretraining for Molecular Force Fields and Properties Prediction

Synthetic pre-training for neural-network interatomic potentials