3.1 Identification of DEGs in ARDS
Raw data of the microarray datasets from GEO datasets were processed by the CEO2R online tool. We extracted 224 and 56 DEGs from GSE130936 and GSE2411, respectively. Then, the common DEGs in the two datasets were identified by Venn diagram software. 39 common DEGs were obtained, including 39 up-regulated genes (log FC> 2, adjust p<0.05) and 0 down-regulated genes(log FC<-2, adjust p<0.05) in the pancreas(Figure 1).
3.2 DEGs gene ontology and KEGG pathway analysis in ARDS
All 39 DEGs were further analyzed by DAVID software. The results of gene ontology analysis showed that 1) For biological process (BP), DEGs were particularly enriched in regulation of immune system process, innate immune response, inflammatory response, cellular response to interferon-beta and so on. 2) For cell component (CC), DEGs were mainly enriched in the extracellular space, the extracellular region, the symbiont-containing vacuole membrane, and the high-density lipoprotein particle. 3) For molecular function (MF), DEGs were enriched in the response to the cytokine activity, the chemokine activity, the CXCR chemokine receptor binding, the chemoattractant activity, the Toll-like receptor 4 binding and the interleukin-1 receptor binding (Figure 2a,2c,2d). The analysis results of KEGG appeared that DEGs were enriched in multiple pathways(Figure 2b), including Cytokine-cytokine receptor interaction, Salmonella infection, Legionellosis, Chemokine signalling pathway, Toll-like receptor signalling pathway and so on (p<0.05).
3.3 Protein-protein interaction network (PPI) and modular analysis
A total of 33 DEGs were imported into the DEGs PPI network complex which included 17 nodes and 77 edges, including 72 up-regulated and 0 down-regulated genes (Figure 3). There were 16 genes excluded from the EDGs PPI network. Then Cytotype MCODE was applied for further analysis. It reviewed that 17 hub genes, including Cd14, Irg1, Iigp1, Gbp6, Ifit1, Ifit2, Saa3, Il1rn, Il1b, Ccl3, Cxcl10, Clec4e, Cxcl1, Cxcl2, Ifit3, Gbp2 and Rsad2, all of which were identified from the 33 nodes. Based on the PPI network analysis, GO term and KEGG pathway enrichment analysis was performed again. The result from GO enrichment analysis showed that hub genes were mostly enriched in the biological process(BP), including cellular response to interferon-beta and so on, in cell components(CC), including symbiont-containing vacuole membrane, extracellular space an extracellular region, and also enriched in the molecular function(MF), including cytokine activity and so on. KEGG pathway was mainly enriched in Salmonella infection signalling pathway and so on (Figure 4). According to the biological process analysis, GBP6, GBP2, IFIT1, IFIT3 and IIGP1 were related to the cellular response to interferon-beta.
3.4 The basic expression of hub genes in the lung and human other organs
The Human Protein Atlas database was used to evaluate the expression level of cor- genes, including GBP6, GBP2, IFIT1, IFIT3 and IIGP1 in varied human organs. IIGP1 was mus musculus specific and not expressed in human. From Figure 5a to 5d, GBP6, GBP2, IFIT1 and IFIT3 were expressed in multiple human organs with different expression levels in different tissues. But GBP6 were not detected in the human pulmonary tissue. It suggested that GBP2, IFIT1 and IFIT3 might be potential targets for ARDS diagnosis and treatment(Figure 5).
3.5 Transcription factor analysis of hub genes
The common transcription factor analysis of the 3 hub genes was conducted using iRegulon, a Cytoscape plugin, and a normalized enrichment shub (NES) >12 was considered to be significant. The transcriptional regulation network of these hub genes was shown in Figure 6. The transcription factors with NES > 12 were STAT1 (NES = 24.252), E2F1 (NES = 21.465), IRF2 (NES = 19.614), IRF1 (NES = 12.027) and IRF9 (NES=12.007).