A novel risk score predicts prognosis in melanoma: The combination of three tumor-infiltrating immune cells and four immune-related genes.

ABSTRACT

Tumor-infiltrating immune cells (TIICs) and immune-related genes (IRGs) of melanoma are associated with prognosis. However, whether the combination of TIICs and IRGs can be used as prognostic clinical biomarkers are still unknown. Here, we downloaded transcription profile of melanoma from TCGA. Then, three TIICs and four IRGs that associated with the overall survival were used to constructed the Immune Cell Score (ICS) and Immune Gene Score (IGS) respectively. Next, to improve the accuracy of ICS and IGS for melanoma prognostic, we combined the ICS and IGS constructed the Immune Cell and Gene Score (ICGS) model. ICGS had higher accuracy and predictive ability than ICS or IGS. Meanwhile, ICGS model reliability was validated by two independent datasets of melanoma. Functional enrichment and protein-protein interaction network analysis based on ICGS were performed to identify T cell mediated immune and inflammatory response are highly associated with melanoma.