Graph-augmented Convolutional Networks on Drug-Drug Interactions Prediction
This addresses drug safety and design for pharmaceutical researchers, though it appears incremental as it combines existing graph CNN and attentive pooling techniques.
The paper tackled drug-drug interaction prediction by proposing an end-to-end model using graph-augmented convolutional networks, achieving high performance with ROC at 0.988, F1-score at 0.956, and AUPR at 0.986.
We propose an end-to-end model to predict drug-drug interactions (DDIs) by employing graph-augmented convolutional networks. And this is implemented by combining graph CNN with an attentive pooling network to extract structural relations between drug pairs and make DDI predictions. The experiment results suggest a desirable performance achieving ROC at 0.988, F1-score at 0.956, and AUPR at 0.986. Besides, the model can tell how the two DDI drugs interact structurally by varying colored atoms. And this may be helpful for drug design during drug discovery.