Molecular Graph Contrastive Learning with Line Graph

TL;DR

LEMON introduces a novel contrastive learning framework that leverages line graphs to better preserve molecular semantics and improve molecular property prediction accuracy.

Contribution

The paper proposes LEMON, a new contrastive learning method using line graphs to enhance molecular semantics encoding and address limitations of existing view generators.

Findings

Outperforms state-of-the-art methods on molecular property prediction

Effectively preserves molecular semantics during contrastive learning

Enhances information transmission with edge attribute fusion and local contrastive losses

Abstract

Trapped by the label scarcity in molecular property prediction and drug design, graph contrastive learning (GCL) came forward. Leading contrastive learning works show two kinds of view generators, that is, random or learnable data corruption and domain knowledge incorporation. While effective, the two ways also lead to molecular semantics altering and limited generalization capability, respectively. To this end, we relate the \textbf{L}in\textbf{E} graph with \textbf{MO}lecular graph co\textbf{N}trastive learning and propose a novel method termed \textit{LEMON}. Specifically, by contrasting the given graph with the corresponding line graph, the graph encoder can freely encode the molecular semantics without omission. Furthermore, we present a new patch with edge attribute fusion and two local contrastive losses enhance information transmission and tackle hard negative samples. Compared with state-of-the-art (SOTA) methods for view generation, superior performance on molecular property prediction suggests the effectiveness of our proposed framework.