Abstract

Info

Title:: Collaborative filtering based on graph neural network for drug-disease association prediction

Author(s):: CHEN Hao1; 2; QIN Yufang1; 2; CHEN Ming1; 2; 1. College of Information Technology, Shanghai Ocean University, Shanghai 201306, China 2. Key Laboratory of Fisheries Information, Ministry of Agriculture and Rural Affairs of the Peoples Republic of China, Shanghai 201306, China

Keywords:: Keywords: drug new indication for drug drug-disease association graph neural network

Abstract:: Abstract: Development?f?ew?rugs takes a long time and is high-cost. Hence its critical to have?fficient and precise methods for predicting new indications for drugs. A drug-disease association prediction method based on graph neural collaborative filtering is proposed in an attempt to obtain information in drug-disease treatment relationships and combine with drug similarities for obtaining better prediction performance. The proposed method firstly capture collaborative signals in drug-disease treatment relationships through graph neural network to refine drug embeddings, then use the drug embeddings to calculate the similarities in drug-disease treatment relationships between drugs, and finally combines with the similarities in drug chemical structures, proteins, and side effects to predict drug repurposing. Compared with the existing collaborative filtering methods on the same data set, the proposed method achieves a higher prediction accuracy (0.664 8). The proposed strategy to obtain potential information in drug-disease treatment relationships and combine with similarities for drug-disease association prediction is effective and helps to discover new indications for drugs and provides assistance in drug development.