Novel Prognostic Factor for Uveal Melanoma: Bioinformatics Analysis of Three Independent Cohorts

Abstract

Background/aim: Because 50% of uveal melanoma metastasize within 10 years of diagnosis, there is urgent need for accurate prognostic factors.

Materials and methods: To identify genes that can act as prognostic factors in uveal melanoma, we performed survival analyses using three independent cohorts. Using log-rank test and univariate cox regression, genes which could stratify the prognosis in all cohorts simultaneously depending on their expression levels were selected as novel biomarkers. Hub genes were obtained by analyzing the interaction and relationship between the selected genes using String and Cytoscape. Additionally, prognostic power was calculated by using C-indices and AUC.

Results: A total of 37 oncogene-like and 14 tumor suppressor-like genes were selected. Protein-protein analysis revealed that NDUFB9, NDUFV2, CYC1 among oncogene-like genes, CTNNB1 among tumor suppressor-like genes were found to be hub genes and core biomarkers in uveal melanoma.

Conclusion: NDUFB9, NDUFV2, CYC1 and CTNNB1 genes may act as prognostic factors in uveal melanoma.

Keywords: Gene Expression Omnibus; The Cancer Genome Atlas; Uveal melanoma; bioinformatics; prognosis.