Chronic obstructive pulmonary disease (COPD) is a heterogeneous and complex syndrome. Network-based analysis implemented by SWIM software can be exploited to identify key molecular switches -called "switch genes" -for disease. Genes contributing to common biological processes or define given cell types are frequently co-regulated and co-expressed, giving rise to expression network modules. Consistently, we found that the COPD correlation network built by SWIM consists of three well-characterized modules: one populated by switch genes, all up-regulated in COPD cases and related to the regulation of immune response, inflammatory response, and hypoxia (like TIMP1, HIF1A, SYK, LY96, BLNK and PRDX4); one populated by well-recognized immune signature genes, all up-regulated in COPD cases; one where the GWAS genes AGER and CAVIN1 are the most representative module genes, both down-regulated in COPD cases. Interestingly, 70% of AGER negative interactors are switch genes including PRDX4, whose activation strongly correlates with the activation of known COPD GWAS interactors SERPINE2, CD79A, and POUF2AF1. These results suggest that SWIM analysis can identify key network modules related to complex diseases like COPD. growth and development, and ineffective lung repair. However, the pathobiological mechanisms for COPD remain incompletely understood [Silverman 2018].COPD susceptibility, like other complex diseases, is rarely caused by a single gene mutation, but is likely influenced by multiple genetic determinants with interconnections between different molecular components. Studying the effects of these interconnections on disease susceptibility could lead to improved understanding of COPD pathogenesis and the identification of new therapeutic targets. Previous efforts to identify the network of interacting genes and proteins in COPD have included protein-protein interaction (PPI) network studies. McDonald and colleagues [McDonald 2014] used dmGWAS software to identify a consensus network module within the PPI network based on COPD GWAS evidence. Sharma and colleagues [Sharma 2018] started with "seed" genes based on well-established COPD GWAS genes or Mendelian syndromes that include COPD as part of the syndrome constellation with a random walk approach to build a COPD network module of 163 proteins.