Fig. 1From: GCMM: graph convolution network based on multimodal attention mechanism for drug repurposingArchitecture of GCMM. a The construction of HN, which contains multi-source drug and disease information. b 1st GCN encoder. It takes HN of drug and disease nodes as input, fuses their neighbor information, and generates embeddings under different views. c 2nd GCN encoder. d Multichannel attention mechanism on drug and disease. e Fully connected feature extractor. f Matrix completion decoderBack to article page