Three-Dimensionally Embedded Graph Convolutional Network (3DGCN) for Molecule Interpretation

Abstract

We present a three-dimensional graph convolutional network (3DGCN), which predicts molecular properties and biochemical activities, based on 3D molecular graph. In the 3DGCN, graph convolution is unified with learning operations on the vector to handle the spatial information from molecular topology. The 3DGCN model exhibits significantly higher performance on various tasks compared with other deep-learning models, and has the ability of generalizing a given conformer to targeted features regardless of its rotations in the 3D space. More significantly, our model also can distinguish the 3D rotations of a molecule and predict the target value, depending upon the rotation degree, in the protein-ligand docking problem, when trained with orientation-dependent datasets. The rotation distinguishability of 3DGCN, along with rotation equivariance, provides a key milestone in the implementation of three-dimensionality to the field of deep-learning chemistry that solves challenging biochemical problems.

arXiv comment39 pages, 14 figures, 5 tables

References in corpus (15)

• Semi-Supervised Classification with Graph Convolutional Networks
• Neural Message Passing for Quantum Chemistry
• SchNet - a deep learning architecture for molecules and materials
• Learning to Plan Chemical Syntheses
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Convolutional Networks on Graphs for Learning Molecular Fingerprints
• AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery
• Neural networks for the prediction organic chemistry reactions
• How Powerful are Graph Neural Networks?
• Hierarchical modeling of molecular energies using a deep neural network
• Modelling Chemical Reasoning to Predict Reactions
• Multi-task Neural Networks for QSAR Predictions
• ChemGAN challenge for drug discovery: can AI reproduce natural chemical diversity?

Cited by in corpus (10)

• Graph Neural Networks: Taxonomy, Advances and Trends
• Molecule Attention Transformer
• Graph Neural Networks for Molecules
• HamNet: Conformation-Guided Molecular Representation with Hamiltonian Neural Networks
• Multitask Learning On Graph Neural Networks Applied To Molecular Property Predictions
• Predicting Material Properties Using a 3D Graph Neural Network with Invariant Local Descriptors
• Deep Multi-attribute Graph Representation Learning on Protein Structures
• InteractionNet: Modeling and Explaining of Noncovalent Protein-Ligand Interactions with Noncovalent Graph Neural Network and Layer-Wise Relevance Propagation
• Distance-Geometric Graph Convolutional Network (DG-GCN) for Three-Dimensional (3D) Graphs
• Geometric Graph Representations and Geometric Graph Convolutions for Deep Learning on Three-Dimensional (3D) Graphs