Abstract
This study addresses the challenges in the early diagnosis of deep infiltrating endometriosis (DIE) by exploring the potential role of the deubiquitinating enzyme USP14. By analyzing the GSE141549 dataset from the Gene Expression Omnibus (GEO) database, using bioinformatics methods and three machine learning algorithms (LASSO, Random Forest, and Support Vector Machine), the key feature gene USP14 was identified. The results indicated that USP14 is significantly upregulated in DIE and exhibits good predictive value (AUC = 0.786). Further analysis revealed the important role of USP14 in muscle function, cellular growth factor response, and maintenance of chromosome structure, and its close association with various immune cell functions. Immunohistochemical staining confirmed the high expression of USP14 in DIE tissues. This study provides a new molecular target for the early diagnosis of DIE, which holds significant clinical implications and potential application value.