Machine learning unveils hypoxia-immune gene hub for clinical stratification of thyroid-associated ophthalmopathy.

ABSTRACT: Thyroid-associated ophthalmopathy (TAO) is an autoimmune disorder affecting the orbit, potentially resulting in blindness. This study focused on the role of hypoxia in its pathogenesis through integrative bioinformatics and experimental validation. Five differentially expressed genes associated with hypoxia (HRDEGs) were identified via Gene Expression Omnibus (GEO) database mining: AGO2, CP, DIO3, PSMD14, WTIP. qPCR and immunohistochemistry confirmed reduced expressions of AGO2 and PSMD14, and elevated expression of DIO3 in TAO orbital tissues. Hypoxia exposure aggravated the above dysregulation and promoted proliferation and adipogenesis of orbital fibroblasts. A predictive model was developed using four machine learning algorithms and validated for its effectiveness in diagnosing TAO and

SUBMITTER: Chen L 

PROVIDER: S-EPMC12214582 | biostudies-literature | 2025 Jul

REPOSITORIES: biostudies-literature