Communicative Representation Learning on Attributed Molecular Graphs

Ying Song, Shuangjia Zheng, Zhangming Niu, Zhang-Hua Fu, Yutong Lu, Yuedong Yang

2020 · 164 citations · 25 references

DOIFull text

Open access

Concepts

TL;DR

Molecular representation is central to property prediction and drug design, and graph neural networks have achieved remarkable success, yet existing MPNN methods focus only on node or edge information, leaving attributed molecular graphs insufficiently represented. The authors propose a Communicative Message Passing Neural Network that enhances molecular embeddings by enabling bidirectional message interactions between nodes and edges via a communicative kernel. The CMPNN strengthens node–edge message interactions with a communicative kernel and further enriches message generation through a booster module. Experiments on six chemical property datasets show that CMPNN outperforms state‑of‑the‑art baselines and visualizations confirm its superior representation capacity.

Abstract

Constructing proper representations of molecules lies at the core of numerous tasks such as molecular property prediction and drug design. Graph neural networks, especially message passing neural network (MPNN) and its variants, have recently made remarkable achievements in molecular graph modeling. Albeit powerful, the one-sided focuses on atom (node) or bond (edge) information of existing MPNN methods lead to the insufficient representations of the attributed molecular graphs. Herein, we propose a Communicative Message Passing Neural Network (CMPNN) to improve the molecular embedding by strengthening the message interactions between nodes and edges through a communicative kernel. In addition, the message generation process is enriched by introducing a new message booster module. Extensive experiments demonstrated that the proposed model obtained superior performances against state-of-the-art baselines on six chemical property datasets. Further visualization also showed better representation capacity of our model.

References

25