Current browse context:
q-bio.QM
Change to browse by:
References & Citations
Quantitative Biology > Quantitative Methods
Title: HyGNN: Drug-Drug Interaction Prediction via Hypergraph Neural Network
(Submitted on 25 Jun 2022 (v1), revised 14 Jul 2022 (this version, v2), latest version 18 Apr 2023 (v4))
Abstract: Drug-Drug Interactions (DDIs) may hamper the functionalities of drugs, and in the worst scenario, they may lead to adverse drug reactions (ADRs). Predicting all DDIs is a challenging and critical problem. Most existing computational models integrate drug-centric information from different sources and leverage them as features in machine learning classifiers to predict DDIs. However, these models have a high chance of failure, especially for the new drugs when all the information is not available. This paper proposes a novel Hypergraph Neural Network (HyGNN) model based on only the SMILES string of drugs, available for any drug, for the DDI prediction problem. To capture the drug similarities, we create a hypergraph from drugs' chemical substructures extracted from the SMILES strings. Then, we develop HyGNN consisting of a novel attention-based hypergraph edge encoder to get the representation of drugs as hyperedges and a decoder to predict the interactions between drug pairs. Furthermore, we conduct extensive experiments to evaluate our model and compare it with several state-of-the-art methods. Experimental results demonstrate that our proposed HyGNN model effectively predicts DDIs and impressively outperforms the baselines with a maximum ROC-AUC and PR-AUC of 97.9% and 98.1%, respectively.
Submission history
From: Khaled Mohammed Saifuddin [view email][v1] Sat, 25 Jun 2022 22:48:27 GMT (938kb,D)
[v2] Thu, 14 Jul 2022 06:31:25 GMT (1567kb,D)
[v3] Fri, 15 Jul 2022 10:14:45 GMT (1567kb,D)
[v4] Tue, 18 Apr 2023 09:57:00 GMT (1568kb,D)
Link back to: arXiv, form interface, contact.