MPHGCL-DDI: Meta-Path-Based Heterogeneous Graph Contrastive Learning for Drug-Drug Interaction Prediction

Abstract

The combinatorial therapy with multiple drugs may lead to unexpected drug-drug interactions (DDIs) and result in adverse reactions to patients. Predicting DDI events can mitigate the potential risks of combinatorial therapy and enhance drug safety. In recent years, deep models based on heterogeneous graph representation learning have attracted widespread interest in DDI event prediction and have yielded satisfactory results, but there is still room for improvement in prediction performance. In this study, we proposed a meta-path-based heterogeneous graph contrastive learning model, MPHGCL-DDI, for DDI event prediction. The model constructs two contrastive views based on meta-paths: an average graph view and an augmented graph view. The former represents that there are connections between drugs, while the latter reveals how the drugs connect with each other. We defined three levels of data augmentation schemes in the augmented graph view and adopted a combination of three losses in the model training phase: multi-relation prediction loss, unsupervised contrastive loss and supervised contrastive loss. Furthermore, the model incorporates indirect drug information, protein-protein interactions (PPIs), to reveal latent relations of drugs. We evaluated MPHGCL-DDI on three different tasks of two datasets. Experimental results demonstrate that MPHGCL-DDI surpasses several state-of-the-art methods in performance.

Keywords: data augmentation; drug-drug interaction; heterogeneous graph contrastive learning; meta-path; protein–protein interaction.

Figure 1

The frequency of DDI events in Dataset1.

Figure 2

Results of MPHGCL-DDI and baselines on events with different frequencies.

Figure 3

Experimental results of MPHGCL-DDI and its five variants in terms of AUPR and macro-F1 on three tasks.

Figure 4

Performance comparison for each DDI event of Dataset1.

Figure 5

Macro-F1 of MPHGCL-DDI with different masking probabilities. (a–f) Dataset1; (g–l) Dataset2.

Figure 6

Performance of MPHGCL-DDI with hyper-parameters $\tau$ and $\alpha$.

Figure 7

The overall framework of the MPHGCL-DDI model.

Figure 8

Illustration of three levels of augmentation schemes.