Abstract

Objective To screen immune-related biomarkers in diagnosing patients with both endometriosis (EM) and systemic lupus erythematosus (SLE). Methods After performing differential expression analysis, immune infiltration analysis, WGCNA, the immune-related genes in EM and SLE were screened. Then the diagnostic genes were identified by three machine learning algorithms, followed by evaluation of the predictive performance of the diagnostic genes by nomogram and ROC curve. Then the correlation between diagnostic genes and immune cells, TFs, GSEA, and potential drugs prediction analyses were performed. Lastly, experiments in vitro were applied to explore the function of SUPT16H in EM and SLE. Results Total 20 immune-related genes in EM and SLE were identified by intersecting DEGs and module genes. Using “LASSO”, “RF”, and “SVM-RFE” algorithms, and total three common diagnostic genes were obtained, namely, C1QC, SOCS3, and SUPT16H. ROC curve shown that AUCs of diagnostic genes were all above 0.7 in training and verification datasets. The targeted drugs for the three diagnostic genes were predicted, containing Pingyangmycin CTD 00001211, VANADIUM PENTOXIDE CTD 00002655, CTD 00001728, and so forth. Also, SUPT16H exerted significant function in occurrence of EM and SLE via regulating inflammation and oxidative stress in cell experiments in vitro. Conclusion SUPT16H overexpression alleviates the progression of EM and SLE by inhibiting inflammation and oxidative stress. The three co- susceptibility genes (C1QC, SOCS3, and SUPT16H) that strongly related to immunity in EM and SLE could be the promising candidate biomarker for the diagnosis and treatment for EM and SLE patients.