Relation-aware Heterogeneous Graph Transformer based drug repurposing

Xin Mei, Xiaoyan Cai, Libin Yang, Nanxin Wang

Abstract: Highlights•A novel neural model, called RHGT, is proposed for drug repurposing.•RHGT characterizes the heterogeneity of the network at node level and edge level.•A fine-grained method is developed to learn edge type embedding.•RHGT achieves state-of-art performance in CTD and TTD datasets.