In this part, the DEseq2 normalized counts of all samples were divided into three groups (blood (PB1 and PB2), intestine (ICV and JJ), and salivary gland (Gland-M and Gland-P)) according to similar tissues, so as to meet the minimum number of samples (n > 10) in each group in this analysis. Then, based on differential expression analysis, the average FDR of each gene was calculated, and 5000 genes with the most significant FDR in each group were selected for the following analysis. Using the WGCNA R package(Langfelder & Horvath, 2008), standard (non_consensus) gene co-expression network analysis in MAP infected samples was performed in each group separately, and the consensus network analysis between MAP infected and healthy samples in each group was then employed to obtain the MAP infected set-specific modules and the consensus modules.