Development of Multiscale Transcriptional Regulatory Network in Esophageal Cancer Based on Integrated Analysis

Objective. To explore multiscale integrated analysis methods in identifying key regulators of esophageal cancer (ESCA). Methods. We downloaded the ESCA dataset from The Cancer Genome Atlas (TCGA) database, which contained RNA-seq data, miRNA-seq data, methylation data, and clinical phenotype information. Then, we combined ESCA-related genes from the NCBI-GENE and OMIM databases and RNA-seq dataset from TCGA to analyze differentially expressed genes (DEGs). Meanwhile, differentially expressed miRNAs (DEmiRNAs) and genes with differential methylation levels were identified. The pivot–module pairs were established using the RAID v2.0 database and TRRUST v2 database. Next, the multifactor-regulated functional network was constructed based on the above information. Additionally, gene corresponding targeted drug information was obtained from the DrugBank database. Moreover, we further screened regulators by assessing their diagnostic value and prognostic value, especially the value of distinguishing patients at TNM I stage from normal patients. In addition, the external database from the Gene Expression Omnibus (GEO) database was used for validation. Lastly, gene set enrichment analysis (GSEA) was performed to explore the potential biological functions of key regulators. Results. Our study indicated that CXCL8, CYP2C8, and E2F1 had excellent diagnostic and prognostic values, which may be potential regulators of ESCA. At the same time, the good early diagnosis ability of the three regulators also provided new insights for the diagnosis and early treatment of ESCA patients. Conclusion. We develop a multiscale integrated analysis and suggest that CXCL8, CYP2C8, and E2F1 are promising regulators with good diagnostic and prognostic values in ESCA.