Figure 4

Cluster analysis demonstrates an altered pattern of transcript expression in the bladders of diabetic and sucrose-treated rats. κ-means clustering grouped transcripts into eight clusters (A) Rows are the mean transcript expression levels (Control n = 6; Diabetic n = 9; Sucrose-treated n = 4) denoted as the z‐score displayed in colourised (High (Red) to Low (Blue) expression) scale. Columns are experimental groups. See Supp. Fig. 7 for full cluster analysis including transcript names. (B) Analysis of the z-score profiles in each cluster. Data are expressed as the mean Cluster z-scores for individual animals in each group (mean ± standard deviation; one-way ANOVA followed by Tukey’s post-hoc test; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (C) The most significant over‐represented KEGG pathway and reactome are shown for each cluster (see Suppl. Fig. 8 for larger image. (D, E) STRING network analysis of Cluster 8 highlights interaction networks and the molecules for the most significant reactome pathway (‘extracellular matrix organisation’) are highlighted as red coloured and annotated nodes.