Edge Attention-based Multi-Relational Graph Convolutional Networks

By Chao Shang, Qinqing Liu, Ko-Shin Chen, Jiangwen Sun, Jin Lu, Jinfeng Yi, Jinbo Bi. 2018

96 citations recorded by the source corpus.

Graph convolutional network (GCN) is generalization of convolutional neural network (CNN) to work with arbitrarily structured graphs. A binary adjacency matrix is commonly used in training a GCN. Recently, the attention mechanism allows the network to learn a dynamic and adaptive aggregation of the neighborhood. We propose a new GCN model on the graphs where edges are characterized in multiple views or precisely in terms of multiple relationships. For instance, in chemical graph theory, compound structures are often represented by the hydrogen-depleted molecular graph where nodes correspond to atoms and edges correspond to chemical bonds. Multiple attributes can be important to characterize chemical bonds, such as atom pair (the types of atoms that a bond connects), aromaticity, and whether a bond is in a ring. The different attributes lead to different graph representations for the same molecule. There is growing interests in both chemistry and machine learning fields to directly learn molecular properties of compounds from the molecular graph, instead of from fingerprints predefined by chemists. The proposed GCN model, which we call edge attention-based multi-relational GCN (EAGCN), jointly learns attention weights and node features in graph convolution. For each bond attribute, a real-valued attention matrix is used to replace the binary adjacency matrix. By designing a dictionary for the edge attention, and forming the attention matrix of each molecule by looking up the dictionary, the EAGCN exploits correspondence between bonds in different molecules. The prediction of compound properties is based on the aggregated node features, which is independent of the varying molecule (graph) size. We demonstrate the efficacy of the EAGCN on multiple chemical datasets: Tox21, HIV, Freesolv, and Lipophilicity, and interpret the resultant attention weights.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Materials Chemistry

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Materials Chemistry

Citations in the corpus

End-to-End Structure-Aware Convolutional Networks for Knowledge Base Completion (2018, unknown): “In contrast, graph convolutional network (GCN) has been an effective tool to create node embeddings which aggregate local information in the graph neighborhood for each node (Kipf and Welling 2016b; Hamilton, Ying, and Leskovec 2017a; Kipf and Welling 2016a; Pham et al. 2017; Shang et al. 2018).”
Spatial Graph Convolutional Networks (2020, methodology): “In addition, EAGCN [13] and MAT [9] are included in the experiment, both of them using an attention mechanism.”
Drug discovery with explainable artificial intelligence (2020, methodology): “64 Attention-based graph convolutional neural networks have also been used for the prediction of solubility, polarity, synthetic accessibility, and photo-voltaic eﬃciency, among other properties, 65,66 leading to the identiﬁcation of relevant molecular substructures for the target properties.”
Graph Warp Module: an Auxiliary Module for Boosting the Power of Graph Neural Networks in Molecular Graph Analysis (2019, methodology): “Many molecular datasets consist of molecular graphs with feature vectors associated to each atom, and numerous methods based on GNN has been proposed to date just for learning the features of chemical molecules [30, 7, 12, 15, 8, 24], such as those pertaining to electrical conductivity and toxicity.”
Deep learning in drug discovery: an integrative review and future challenges (2022, background): “suggested a multi-relational GCN with edge attention (Shang et al. 2018).”
Graph convolutional networks for computational drug development and discovery (2019, background): “[36] proposed edge attention-based graph convolution network to deal with multi-relational graphs.”