Background: Melanoma is a cancer of the skin with potential to spread to other organs and is responsible for most deaths due to skin cancer. It is imperative to identify immune biomarkers for early melanoma diagnosis and treatment.

Results: 63 immune-related genes of the total 1039 unique IRGs retrieved were associated with overall survival of melanoma. A multi-IRGs classifier constructed using eight IRGs showed a powerful predictive ability. The classifier had better predictive power compared with the current clinical data. GSEA analysis showed multiple signaling differences between high and low risk score group. Furthermore, biomarker was associated with multiple immune cells and immune infiltration in tumor microenvironment.

Conclusions: The immune-related genes prognosis biomarker is an effective potential prognostic classifier in the immunotherapies and surveillance of melanoma.

Methods: Melanoma samples of genes were retrieved from TCGA and GEO databases while the immune-related genes (IRGs) were retrieved from the ImmPort database. WGCNA, Cox regression analysis and LASSO analysis were used to classify melanoma prognosis. ESTIMATE and CIBERSORT algorithms were used to explore the relationship between risk score and tumor immune microenvironment. GSEA analysis was performed to explore the biological signaling pathway.