Hierarchical clustering
In order to identify differences and similarities among the different samples analyzed here (SLIL zones and controls) based on their global histochemical and immunohistochemical profiles, we analyzed all samples using hierarchical clustering. With this purpose, average results obtained for all quantitative histochemical and immunohistochemical analyses were analyzed using ClustVis web tool for clustering of multivariate data (Metsalu & Vilo, 2015) (accessible athttps://biit.cs.ut.ee/clustvis/). A Heatmap dendrogram displaying relative quantitative signals for each marker was generated by the software based on hierarchical clustering analysis using the Euclidean distance and the complete clustering method for both the rows and columns.