DapBCH: a disease association prediction model Based on Cross-species and Heterogeneous graph embedding

The study of comorbidity can provide new insights into the pathogenesis disease and has important economic significance in clinical evaluation treatment difficulty, medical expenses, length stay, prognosis disease. In this paper, we propose a association prediction model DapBCH, which constructs cross-species biological network applies heterogeneous graph embedding to predict association. First, combine human disease–gene network, mouse gene–phenotype human–mouse homologous gene protein–protein interaction reconstruct network. Second, apply based on meta-path aggregation generate feature vector nodes. Finally, employ link obtain similarity pairs. experimental results indicate that our is highly competitive predicting promising for finding potential associations.