I have 32 samples (16 +16 of two conditions) of RNASeq data and after differential expression analysis using DESeq2, I used WGCNA to make gene modules. To visualize the networks to which the module genes belong, I used ReactomePA to perform enrichment analysis of all genes in each module. I observe that the top genes with highest "kwithin" and "ktotal" do not show up in the reactomePA enriched gene networks (cnetplot).