Identification of human microRNA-disease association via hypergraph embedded bipartite local model. (December 2020)

• Record Type: Journal Article
• Authors: Ding, Yijie; Jiang, Limin; Tang, Jijun; Guo, Fei
• Is Part Of: Computational biology and chemistry. Volume 89(2020)
• Publication Date: 2020-12
• Subjects: Human MicroRNA-disease association -- Bipartite network -- Hypergraph learning -- Laplacian support vector machine -- Graph regularized model; Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2020.107369
• Languages: English
• ISSNs: 1476-9271

Abstract

Graphical abstract: Flow diagram of hypergraph regularized bipartite local model (HGBLM). Highlights: HGBLM approach builds more complex relationships between samples by hypergraph. Heterogeneous information (similarity matrices) is integrated via CKA-MKL. HGBLM is built via hypergraph embedded LapSVM to further improve the performance of prediction. Abstract: MicroRNA (miRNA) plays an important role in life processes. In recent years, predicting the association between miRNAs and diseases has become a research hotspot. However, biological experiments take a lot of time and cost to identify pathogenic miRNAs. Computational biology-based methods can effectively improve accuracy of recognition. In our study, miRNAs-disease associations are predicted by a hypergraph regularized bipartite local model (HGBLM), which is based on hypergraph embedded Laplacian support vector machine (LapSVM). On benchmark dataset, the results of our method are comparable and even better than existing models.