Fig. 3

Weighted gene co-expression network analysis of lipid metabolism- and immune-related genes. A Determination of soft-threshold power in WGCNA. The scale-free topology index and the mean connectivity for each power value between 1 and 30 were shown. The appropriate soft-thresholding power was picked for network construction. B Validation of the optimal soft threshold power by the high scale-free topology R² between log₁₀(k) and log₁₀(p(k)). k represents the connectivity between genes; and p(k) represents the probability of connectivity. C Hierarchical clustering tree based on WGCNA module eigengenes. Five modules were identified including grey, yellow, brown, blue, and turquoise modules. D The hierarchical clustering dendrogram of the eigengenes based on hierarchical clustering under optimal soft-thresholding power. Dynamic tree cut represents module divided according to clustering results; and merged dynamic represents module divided according to similarity of the module. E Bar plot of mean gene significance across various modules. F The bar plot showing the gene counts in different modules. The pink bar represents the immune-related genes; the emerald green bar represents the lipid metabolism-related genes; and the blue bar represents genes related to both immune and lipid metabolism