GENN: Predicting Correlated Drug-drug Interactions with Graph Energy Neural Networks
This work addresses the need for more accurate DDI predictions in drug development and medical practice, though it is incremental by focusing on link type correlations.
The paper tackled the problem of predicting drug-drug interactions (DDIs) by modeling correlations between link types, which were often ignored in prior graph neural network approaches, and achieved improvements of 13.77% and 5.01% in PR-AUC on two datasets.
Gaining more comprehensive knowledge about drug-drug interactions (DDIs) is one of the most important tasks in drug development and medical practice. Recently graph neural networks have achieved great success in this task by modeling drugs as nodes and drug-drug interactions as links and casting DDI predictions as link prediction problems. However, correlations between link labels (e.g., DDI types) were rarely considered in existing works. We propose the graph energy neural network (GENN) to explicitly model link type correlations. We formulate the DDI prediction task as a structure prediction problem and introduce a new energy-based model where the energy function is defined by graph neural networks. Experiments on two real-world DDI datasets demonstrated that GENN is superior to many baselines without consideration of link type correlations and achieved $13.77\%$ and $5.01\%$ PR-AUC improvement on the two datasets, respectively. We also present a case study in which \mname can better capture meaningful DDI correlations compared with baseline models.