Abstract
Molecular pathways regulating the initiation and development of melanoma are potential therapeutic targets for this aggressive skin cancer. Therefore, transcriptome profiles of cutaneous melanoma were obtained from a public database and used to systematically evaluate cancer hallmark pathways enriched in melanoma. Finally, the unfolded protein response pathway was screened out, and the unfolded protein response-related genes were used to develop a robust biomarker that can predict the prognosis of melanoma, especially for younger, metastatic and high Clark level patients. This biomarker was further validated in two other independent datasets. In addition, melanoma patients were divided into high- and low-risk subgroups by applying a risk score system. The high-risk group exhibited higher immune infiltration and higher expression of N6-methyladenosine RNA methylation regulators, and had significantly shorter survival times than the low-risk subgroup. Gene Set Enrichment Analysis revealed that, among the enriched genes, gene sets involved in immune response and the extracellular matrix receptor interaction were significantly activated in the high-risk group. Our findings thus provide a new clinical application for prognostic prediction as well as potential targets for treatment of melanoma.