Disease and Gene Association Mining Based on Graph Data Enhancement
In view of the incompleteness of existing association data and the inadequacy of multi-source omics data,computational indexes based on three-hop local topological similarity were designed to identify biologically significant but unmapped Protein-Protein Interactions(PPI).A novel graph neural network method(GDaEPred)based on graph data enhancement was proposed for mining disease-gene associations.Experimental results showed that the average accuracy of GDaEPred was improved by 4.1%,and the pre-cision,recall and F1 score were also improved.
graph neural networksgraph data enhancementdisease gene prediction
国家科技期刊平台
NSTL
万方数据
维普
