Fig. 2

Identification of key modules highly correlated with stage I to IV in the TCGA COAD-READ dataset. A The sample clustering tree of CRC samples in the TCGA dataset to find and remove outlier samples. B Analysis of scale-free fit index (left) and mean connectivity (right) for each soft-threshold power. C Clustering of module eigengenes with the red line representing cut height (0.4). D Clustering dendrograms of robust DEGs and related modules based on a dissimilarity measure (1- TOM). E Heatmap of the correlation between module eigengenes and stage I to IV (as clinical traits) in CRC. Each cell contains p-value and the correlation coefficient. Scatter plot of key modules related to stage I (F), stage II (G), stage III (H) and stage IV (I)