Drug Repositioning with GraphSAGE and Clustering Constraints Based on Drug and Disease Networks

The study introduced a drug repositioning method called DRGCC, which utilized GraphSAGE and clustering constraints to predict drug-disease associations by reconstructing drug interaction and disease similarity networks. It demonstrated superior performance compared to six state-of-the-art methods, achieving an AUC of 0.9809 on the CTD dataset. The method was applied to anti-COVID-19 drug prediction, identifying several potential drugs, such as triazavirin and posaconazole, supported by molecular docking analysis and existing research. The study analyzed 780 drugs, 717 diseases, and 17,594 therapeutic associations, showing the method's potential to rapidly and accurately discover drug candidates for viral infectious diseases.