RNA-seq data and clinical features of melanoma samples were obtained from The Cancer Genome Atlas. Samples of melanoma were randomly assigned to the training and testing cohort. The immune-related lncRNA signature was then obtained via using univariate, LASSO, and multivariate Cox analysis of patients in the training cohort. Eight significant immune-related lncRNA signature was then subsequently obtained through correlation analysis between immune-related genes and lncRNAs. The association between risk score and immune cell infiltration was finally assessed using TIMER and CIBERSORT.
Three hundred and fifty-six immune-related lncRNAs were obtained. Among them, eight immune-related lncRNAs were identified to build a prognostic risk signature model. The model’s performance was then confirmed using the Kaplan-Meier curves, risk plots, and time-dependent receiver-operating characteristic curves in the training cohort. The risk score was identified and confirmed as an independent prognostic factor through univariate and multivariate Cox regression analyses. These results were further verified in the testing and whole cohorts. CIBERSORT algorithm showed that the infiltration levels of T cells CD8, M1 macrophages, plasma cells, T cells CD4 memory activated, T cells gamma delta, and mast cells activated were significantly lower in the high-risk group while the infiltration level of macrophages M0 was significantly lower in the low-risk group.
The immune-related lncRNA signature offers prognostic markers and potential immunotherapeutic targets for melanoma.
© 2021 The Authors. Immunity, Inflammation and Disease published by John Wiley & Sons Ltd.