189. IDENTIFICATION OF M6A-RELATED SIGNATURE GENES IN ESOPHAGEAL SQUAMOUS CELL CARCINOMA BY MACHINE LEARNING METHOD

Abstract Background We aimed to construct and validate the esophageal squamous cell carcinoma (ESCC)-related m6A regulators by means of machine leaning. Methods used ESCC RNA-seq data 66 pairs from West China Hospital Sichuan University transcriptome extracted The Cancer Genome Atlas (TCGA)-ESCA database find out ESCC-related regulators, during which, two learning approaches: RF (Random Forest) SVM (Support Vector Machine) were employed model regulators. Calibration curves, clinical decision impact curves (CIC) evaluate predictive ability best-effort model. Finally, western blot immunohistochemistry staining assess expression prognostic Results 2 (YTHDF1 HNRNPC) found be significantly increased in tissues after screening through methods our TCGA-ESCA database, respectively, overlapping results clusters. A signature, consisting YTHDF1 HNRNPC, was constructed based on validated which can serve as an independent predictor. Experimental validation including further successfully confirmed bioinformatics analysis. Conclusion cohort well tissues, provides reasonable evidence valuable resources for stratification study potential targets ESCC.