Figure 1.

Hierarchical cluster trees for detecting modules based on weighted gene co-expression network analysis. All array samples (primary Sjögren's syndrome patients, sicca, and controls) were to find co-expression modules. Modules correspond to branches of the trees and are assigned color categories as indicated by the color band underneath each tree (see Materials and methods). The two panels correspond to two cluster trees resulting from the blockwiseModules WGCNA R function [21], which was used to circumvent computational challenges. Each cluster tree thus corresponds to a block of genes.