Skip to main content
Fig. 1 | BMC Bioinformatics

Fig. 1

From: Drug-target interaction prediction using semi-bipartite graph model and deep learning

Fig. 1

Our model workflow. a Data representation. A semi-bipartite graph is constructed by Drug-target interactions, drug-drug similarity and protein-protein similarities. b Drug-target positive and negative pairs (extracted based on [29]) samples are represented as sub-graphs capturing the topological environment around drug target pairs. c A graph labelling method is applied on each sub-graph in order to preserve the ordering of graph vertices. d The final sub-graphs are converted to adjacency matrices and the upper triangle of each matrix is representing embedded features to train a classifier. e A deep neural network is trained and used for predicting new drug target pairs

Back to article page