Fig. 4

a Protein–protein interaction network of 65 genes imported from the String database and hiding disconnection points. b The genes with the top ten Degree values were selected as Hubs according to CytoHubba's built-in algorithm. c The correlation heatmap of the ten hub genes. The numbers in the squares running across the diagonal represent the EntrezID of the gene in the heat map. d Gene–gene interaction network of ten hub genes, showing 20 most frequently changing neighboring genes. Each node represents a gene. The color of the edges reflects the roles of physical interactions, co-expression, prediction, pathway, co-localization, genetic interactions, and shared protein domains, respectively. e–h Enrichment results of 10 Hub genes in GO functional annotation (BP, CC, and MF) and KEGG signaling pathway enrichment results