Weighted Gene Correlation Network Analysis (WGCNA)
Groups with the greatest genetic distance and DEGs were selected for WGCNA. The analysis was conducted using the WGCNA R package following the provided tutorials (Langfelder & Horvath, 2008). The soft thresholding power was determined according to the principle of a nonscale network, and the lowest power when the correlation coefficient reached the plateau period was used as a parameter in subsequent analysis. A gene clustering tree was constructed according to the correlations between the expression levels of genes and coexpression modules were identified by dynamic tree cutting using a minimum module size of 30. When the correlation between the modules and traits was greater than 0.75 (p < 0.05), the modules were significantly related to the traits. The R package clusterProfiler was used for KEGG enrichment analysis of genes in the phenotype-related modules (Yu et al., 2012).