To identify the genes that are central and highly connected in the docetaxel-resistance-associated modules, we selected Hub genes with |GS| > 0.4 (gene significance) and |MM| > 0.8 (module membership) in the two modules and visualized these genes by Cytoscape (Fig. 3C).
To investigate the biological functions of the WGCNA modules, we selected the highly connected "hub genes" within each module.
From these modules we selected 2 modules from which the co-expression of the genes was observed in more than 5 samples (to increase the robustness of the analysis).
For the Blue and Yellow modules, we selected gene lists of 500 genes with module membership p-values smaller than 10-22 and corresponding local false discovery rates (q-values) smaller than 10-5.
For each miRNA module, we randomly selected 1000 modules with the same number of miRNAs, calculated the proportion of age-related miRNAs in modules and evaluated the correlation between the expression levels of the modules and age.
To detect LOH regions, we first counted the number of SNVs in 100-kb sequence blocks by VCFtools SNPdensity module, and then we selected those blocks where the number of homozygous SNVs was at least 10-times greater than the number of heterozygous SNVs and the number of homozygous SNVs was more than 50.
On the basis of those selection criteria we selected five modules for functional annotation: the Blue Module (275 genes), the Brown Module (62 genes), the Lightyellow Module (21 genes), the Black Module (105 genes) and the Greenyellow Module (41 genes).
We selected modules detected by SA algorithm for similarity analysis and functional classification.
We selected modules defined at branch lengths 0f 0.8, 0.8, 1.1, 1.9, 2.5, 4.5, 10.6, 24.7 and 101.6 for further analysis since these gave a broad range of cluster numbers.
Using this information, we selected the module shown in Figure 4 for the cdc15 data set.
