identification of piRNA-disease associations based on Supplementarily Weighted Graph Convolutional Network

Home | Download | Citation


Motivation: Accurately identifying potential piRNA-disease associations is of great importance in uncovering the pathogenesis of diseases. Recently, several machine-learning-based methods have been proposed for piRNA-disease association detection. However, they are suffering from the high sparsity of piRNA-disease association network and the Boolean representation of piRNA-disease associations ignoring the confidence coefficients.
Results: In this study, we propose a supplementarily weighted strategy to solve these disadvantages. Combined with Graph Convolutional Networks (GCNs), a novel predictor named iPiDA-SWGCN is proposed for piRNA-disease association prediction. There are three main contributions of iPiDA-SWGCN: (i) Potential piRNA-disease associations are preliminarily supplemented in the sparsity piRNA-disease network by integrating various basic predictors to enrich network structure information. (ii) The original Boolean piRNA-disease associations are assigned with different relevance confidence to learn node representations from neighbour nodes in varying degrees. (iii) The experimental results show that iPiDA-SWGCN is superior to the other state-of-the-art methods, and can predict new piRNA-disease associations. The flowchart of iPiDA-SWGCN model is shown in Fig.1

iPiDA-GCN web server
Fig.1 The framework of iPiDA-SWGCN (a) Network construction. A heterogeneous piRNA-disease network is constructed by integrating piRNA and disease information. (b) Supplementarily weighted piRNA-disease network generation. Different supplementary weights are assigned to unknown piRNA-disease pairs based on the scores computed by several predictors; (c) GCN-based feature extraction. With supplementarily weighted piRNA-disease network, GCN is performed on the heterogeneous piRNA-disease network to capture the structural information, and extract features of piRNAs and diseases. (d) Association prediction. Finally, we use the fully connected layers for dimension reduction and inner product operation so as to calculate the association scores between piRNAs and diseases.