Identification of the susceptible genes of systemic lupus erythematosus and sepsis: based on immune and oxidative stress-related genes

Abstract

Abstract Background: Systemic lupus erythematosus (SLE) is an autoimmune inflammatory connective tissue disease involving multiple organs. As one of the serious complications of SLE, sepsis (SEP) has a high risk of death. Here, the goal of this study was to identify vulnerable biomarkers that could be used to diagnose SLE and SEP. Methods: We used the Limma R software tool and the Gene Expression Omnibus (GEO) database to find differentially expressed genes (DEGs) in SLE. Additionally, genes associated with oxidative stress and immune system function were chosen from the MSigDB database and the Genecard database, respectively. Weighted gene Coexpression network analysis (WGCNA) was used to identify the important module genes associated with SEP. With the help of WGCNA, machine learning, and logistic regression, immunological and oxidative stress-related hub genes were discovered and validated by an external validation set. The analysis was put to the test using consensus clustering. Immune cell infiltration was investigated in SLE and SEP patients. Results: We obtained 957 genes from the GSE6163 dataset and 2559 genes from the significant module of WGCNA, which yielded 46 genes after taking intersection with immune and oxidative stress-related genes. According to the enrichment analysis's findings, the two diseases share a lot of similar immunological and inflammation-related pathways. Machine learning was utilized to pick 11 hub genes, and ROC was employed to evaluate the diagnostic effectiveness. Furthermore, the expression profiles of the hub genes revealed by logistic regression modeling have a significant diagnostic value. Moreover, consensus clustering revealed a favorable correlation between the severity of immunological and oxidative stress and disease activity in SLE and SEP. Analysis of immune infiltration revealed a more consistent immune cell infiltration behavior between SLE and SEP. Conclusion: In this study, the expression of 11 potential hub genes, including TLR2, IL1RN, IRF9, ISG20, TXK, SH2D1A, IL7R, CD28, ITK, CD3E, and CCR7, were thoroughly analyzed using bioinformatics. An efficient logistic regression model was created, and it was possible to identify a correlation between the progression of SLE and SEP disease and the expression of immunological and oxidative stress by consensus clustering. In addition, there is a similar immunoinvasive behavior between the two diseases. It is helpful to identify the biological markers with potential diagnostic value.