Exploring Prognostic Gene Expression Signature for Thyroid Eye Disease

Background: Thyroid-related eye disease (TED) is the most prevalent orbital disorder of autoimmune origin, yet research on its mitochondrial energy metabolism-related genes remains limited with no definitive biomarkers or therapeutic targets identified.Methods: We downloaded datasets from the GEO database, forming TED and Control groups. Mitochondrial energy metabolism-related genes (MEMRGs) were sourced from GeneCards. We conducted differential gene expression analysis, gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment, gene set enrichment analysis (GSEA), and identified Key Genes. A diagnostic logistic regression model was then built, expression differences validated, immune infiltration analyzed via ssGSEA, and network interaction assessed. Results: After correcting batch effects in datasets GSE58331 and GSE105149, we identified 26 mitochondrial energy metabolism-related differentially expressed genes (MEMRDEGs). Through one-way logistic analysis, Random Forest, and a LASSO risk model, we pinpointed 10 Key Genes crucial for diagnosis. The diagnostic model demonstrated superior performance and accuracy, with Key Genes showing significantly higher expression in the TED group. Additionally, activated CD8 T cells, immature B cells, and T follicular helper cells exhibited notable differences. Conclusions: The expression of mitochondrial energy metabolism-related genes in TED, stratified by risk score levels, exhibits a positive correlation with immune activity, unveiling a promising new pathway for identifying therapeutic targets. This finding calls for deeper biological exploration.