*2.4. WGCNA Analysis*

### 2.4.1. Construction of Gene Co-Expression Network of Dendrobium

Consequently, we used the WGCNA package tool to construct a co-expression module that expressed the expression of 8056 genes in 27 *Dendrobium* samples. The heat map of the cluster dendrogram and tissues and species traits are shown in Figure 7a. One of the most critical parameters is the power value, which mainly affects the independence and average connectivity of the co-expression modules. First, we selected the appropriate power value. When the power value was 18, the independence reached 0.6, and the average connectivity was higher. Therefore, the power values and results of constructing co-expression modules indicate that 19 different co-expression modules were identified in *Dendrobium*. These co-expression modules were constructed and are shown in different colors (Figure 7). The size of these modules depends on the number of genes they contain. The number of genes and module names are shown in Table 1.

**Figure 7.** Clustering dendrogram. (**a**) Clustering dendrogram of 27 samples and heatmaps of species and tissues traits. The expression is from low to high, and the color transitions from white to red. (**b**) Clustering dendrogram of DEGs, with dissimilarity based on the topological overlap, together with assigned module colors. The clustered branches represent different modules, and each line represents one gene.


**Table 1.** The number of genes in 19 constructed modules.

2.4.2. Interaction Analysis of Co-Expression Module

Subsequently, we analyzed the interactions between the 19 co-expression modules (Figure 8). The heatmap shows the topological overlap matrix (TOM) of all genes in the analysis. Light color indicates low overlap, and dark red indicates high overlap. Except for some high-brightness areas, the overall difference between the modules is not significant, which indicates that the gene expression between the modules is relatively independent and has a high scale independence. The correlations between module eigengene and traits: tissues and species, were analyzed and data are shown in Figure 8b. Two modules are significantly associated with species: Lightgreen (*p*-value <sup>=</sup> <sup>1</sup> <sup>×</sup> <sup>10</sup>−5, cor <sup>=</sup> 0.73) and Lightsteelblue1 (*p*-value <sup>=</sup> <sup>2</sup> <sup>×</sup> <sup>10</sup>−6, cor <sup>=</sup> <sup>−</sup>0.76). The modules that highly correlated to tissues were Yellowgreen (*p*-value <sup>=</sup> <sup>1</sup> <sup>×</sup> <sup>10</sup>−6, cor <sup>=</sup> 0.79), Salmon4 (*p*-value <sup>=</sup> <sup>9</sup> <sup>×</sup> <sup>10</sup>−7, cor <sup>=</sup> 0.79), Blue (*p*-value <sup>=</sup> <sup>3</sup> <sup>×</sup> <sup>10</sup>−6, cor <sup>=</sup> 0.77), and Sienna3 (*p*-value <sup>=</sup><sup>2</sup> <sup>×</sup> <sup>10</sup>−10, cor <sup>=</sup> 0.9).

We analyzed the connectivity of eigengenes to find the interactions between these constructed co-expression modules. First of all, cluster analysis of these eigengenes was performed (Figure 9a). These 19 clusters were divided into two clusters, including 5 modules (modules 3, 8, 9, 11, and 14) and the remaining 14 modules. According to that, the connectivity effect between different modules is obviously different. (Figure 9b).

**Figure 8.** Co-expression network analysis across different tissues and different species. (**a**) Visualizing the gene network using a heatmap plot. The heatmap depicts the topological overlap matrix (TOM) among all genes in the analysis. (**b**) Module-trait associations. Each row corresponds to a module characteristic gene (eigengene), and each column corresponds to a trait. Each cell contains a corresponding correlation and *p*-value. According to the color legend, the table is color-coded by correlation.

**Figure 9.** Analysis of connectivity of eigengenes in different module. (**a**) Cluster analysis of eigengenes. (**b**) The heatmap of connectivity of eigengenes.
