Fig. 5

Flowchart of genotype-phenotype association based on data integration algorithm. The steps of this method in this figure can be described as follows: 1) PPI network and gene expression data are used to generate a weighted undirected gene association network, which is clustered by SPICi to generate gene clusters; 2) Gene clusters are screened by group lasso; 3) SNP clusters corresponding to the selected gene clusters are obtained from eQTL data; 4) SNP clusters, correlated gene clusters, and phenotype are combined in a three-layer network block, which is used for prediction and analysis through three-layer associations; 5) Results from various blocks are integrated by taking the average to produce the final output