Fig. 2

Establishing correlation patterns among differentially expressed genes (n = 2607) in ovarian cancer using Weighted Gene Co-expression Network (WGCNA) analysis. a Selection of parameter beta for the power transformation of the correlation matrix into the adjacency matrix in order to obtain scale free topology. The horizontal and vertical axes on the left plot represent the soft threshold power and the scale free fit index, respectively. The red line represents the standard line when R^2 reached a value of 0.88. The mean connectivity is depicted as a function of the soft threshold powers. b Similar modules with correlating module eigengenes were merged to form four major modules based on a distance threshold cut-off of 0.25. c Hierarchical clustering of genes with dissimilarity based on topological overlap are shown along with the modules detected and the merged modules. All genes that do not fall in any of the modules are kept in grey module. d Module trait relationship between cancer and normal tissues was established. Each module shows correlation coefficient and corresponding p-value for the correlation of genes in a specific module to the selected traits i.e. cancer or normal tissue. Cyan module showed maximum correlation and was the most relevant module with cancer traits. e Gene significance for ovarian cancer versus the module membership in the cyan module is depicted as a scatter plot. Intramodular analysis of the genes found in the cyan module, which contains genes that have a high correlation with ovarian cancer, with p < 2e-32 and correlation = 0.94