Consensus analysis via weighted gene co-expression network analysis (WGCNA) reveals genes participating in early phase of acute respiratory distress syndrome (ARDS) induced by sepsis

The understanding of mechanism during conversion from sepsis to sepsis-related ARDS remains limited. In this study, we collected gene expression matrix from the Gene Expression Omnibus (GEO) database and constructed networks using weighted gene co-expression network analysis (WGCNA) to identify the consensus and opposite modules between sepsis and sepsis-induced ARDS and obtained 27 consensus modules associated with sepsis and sepsis-related ARDS, including one model (160 genes) with opposite correlation and 1 sepsis-ARDS specific model with 34 genes. Differentially expressed genes analysis, functional enrichment and protein-protein interactions analyses of candidate genes were performed; 15 of these genes showed different expressions in sepsis-induced ARDS patients, compared with sepsis patients; genes were enriched in processes associated with ribosome, tissue mechanics and extracellular matrix. Feature selection analysis revealed that three genes, TLCD4, PRSS30P, and ZNF493, showed moderate performance in identifying sepsis-induced ARDS from sepsis. Ribosome-related genes indicate crucial roles in the development of sepsis-induced ARDS.