Abstract
PURPOSE: Endometriosis (EMS) is a relatively common gynecological disorder and almost fifty percent of women with EMS suffer from infertility. There are few treatment options for endometriosis, and often recurrences occur following surgery and medication. We aimed to identify potential diagnostic biomarkers for EMS to improve its diagnostic efficiency. METHODS: Differential analysis was utilized to choose EMS-associated abnormal miRNAs (DEMIs) and mRNAs (DEMs). ImmuneAI analysis was to evaluate the levels of immune cells in EMS. Next, the weighted gene co-expression network analysis (WGCNA) was utilized to identify the co-expression modules. Random forest and SVM analyses were used to filter the candidate biomarkers and construct the diagnostic model. qRT-PCR was used to test the expression level of the biomarkers. RESULTS: Based on the different analyses, we obtained 32 DEMIs and 516 DEMs and selected 9 abnormal immune cells whose abundance is abnormal in EMS. Next, we identified five co-expression modules associated with these abnormal immune cells. Then, 176 candidate genes which are both miRNA targets and associated with immune cells and aberrantly expressed in EMS were filtered. Subsequently, random forest analysis selected 11 genes as the diagnostic biomarkers and constructed a diagnostic model by SVM. Finally, we demonstrated that 8 of the 11 genes aberrantly expressed and with better diagnostic efficiency in EMS. CONCLUSIONS: In total, we identified 11 crucial genes regulated by 8 miRNAs that could serve as promising diagnostic biomarkers for EMS, potentially enhancing disease diagnosis with novel factors.