Figure 4. WGCNA discoveries the key ANXA1-entered module. (A) Hierarchical clustering dendrogram of tumor samples from TARGET_OS with the indicated clinical traits. (B–C) Scale independence and mean connectivity analyses of OS samples. Soft-thresholding power β = 3 fits the scale-free topology. (D) Clustering tree of genes with divergence. The branches of tree represent different modules. (E) Heatmap of TOM among 1000 genes which were selected randomly in WGCNA. Dark/light color corresponds to the degree of overlap. (F) Relevance of module eigengenes with traits. Based on the correlation and P-value in each cell, the magenta module containing ANXA1 is selected.