Integrated Transcriptomic and Network Analysis Identifies Diagnostic and Immune-Associated Hub Genes in Urothelial Carcinoma

Background: Urothelial carcinoma exhibits molecular heterogeneity, high recurrence, and progression rates, and is poorly diagnosed and treated. Transcriptomic and systems-biology analyses can detect dysregulated genes, signaling pathways, and interactions of immune cells. Proposed biomarkers, however, have not been validated, evaluated, and confirmed to be cell-type specific, limiting their potential for clinical translation. Methods: The transcriptomic data in GSE13507 were used to identify features of differential expression, functional enrichment, protein interaction network, and hub genes. Findings were assessed by using external validation, ROC and machine learning models, immune-infiltration analysis, single-cell expression profiling, and survival analysis. Results: In total, 7459 differentially expressed probes were detected, and pathways involved in immunity and cytokines were noted to be enriched. The best diagnostic performance was observed with IL6ST (AUC 0.927), LIFR (AUC 0.921), and JAK1 (AUC 0.843). We found a correlation between hub genes and infiltration of immune cells, and a difference in single-cell distribution; however, none of them was shown to have a significant association with overall survival. Conclusions: The IL-6/JAK/STAT hub gene network was highly diagnostic and immune-relevant in urothelial carcinoma and had limited prognostic significance. The three most promising candidates were IL6ST, LIFR, and JAK1. The genes identified in this study could be validated in tumor-normal pairs and in prospective clinical samples of future studies, accompanied by protein levels and functional studies.