Identification of MAPK12 as a Prognostic Biomarker for Esophageal Carcinoma Using Bioinformatics and Machine Learning

To develop a telomere-related prognostic signature for esophageal carcinoma (ESCA), we integrated bioinformatics and machine learning approaches. Hub genes were identified from overlapping differentially expressed genes (DEGs). A prognostic model was constructed using LASSO and multivariate Cox regression, validated in independent GEO datasets, and further verified through cytological experiments. We also elucidated the mechanism by which MAPK12 promotes ESCA migration. The model robustly predicted survival of patients with ESCA, supported by both high-throughput data and experimental evidence. Our findings highlight MAPK12 as a promising biomarker and provide a theoretical basis for understanding ESCA pathogenesis and developing targeted therapies.