A prognostic model for melanoma patients on the basis of immune-related lncRNAs

Abstract

The prognosis of melanoma patients is highly variable due to multiple factors conditioning immune response and driving metastatic progression. In this study, we have correlated the expression of immune-related lncRNAs with patient survival, developed a prognostic model, and investigated the characteristics of immune response in the diverse groups. The gene expression profiles and prognostic information of 470 melanoma patients were downloaded from TCGA database. Significantly predictive lncRNAs were identified by multivariate Cox regression analyses, and a prognostic model based on these variables was constructed to predict survival. Kaplan-Meier curves were plotted to estimate overall survival. The predictive accuracy of the model was evaluated by the area under the ROC curve (AUC). Principal component analysis was used to observe the distribution of immune-related genes. CIBERSORT and ESTIMATE were used to evaluate the composition of immune cells and the immune microenvironment. Eight immune-related lncRNAs were determined to be prognostic by multivariate COX regression analysis. The patient scores were calculated and divided into high- and low-risk groups. The model could effectively predict the prognosis in patients of different stages. The AUC of the model is 0.784, which was significantly higher than that of the other variables. There were significant differences in the distribution of immune-related genes between two groups; the immune score and immune function enrichment score were higher in the low risk group.

Construction of a novel risk model based on the random forest algorithm to distinguish pancreatic cancers with different prognoses and immune microenvironment features

“…Immune-related lncRNAs (irlncRNAs) have recently been studied in several cancers. Wang et al [ 9 ] identified 4 irlncRNAs to establish a risk model for lung adenocarcinoma, while 8 irlncRNAs were used to construct a prognostic model for melanoma [ 10 ]. Additionally, irlncRNAs have been utilized in glioblastoma, head and neck squamous cell carcinoma and bladder cancer [ 11–13 ].…”

The Functional Role of Long Non-Coding RNAs in Melanoma

“…Several reports show that various lncRNAs, such as XIST, NEAT1 or NKILA, are generally involved in the regulation of human immune system and tumor-immune cell infiltration [ 215 , 216 , 217 ]. The SKCM dataset from TCGA, that comprises molecular and clinical data for 470 melanoma patients has been used recently to identify immune-related lncRNAs [ 218 , 219 ]. Ping et al have applied a modified least absolute shrinkage and selection operator (LASSO) regression model to identify 28 immune-related lncRNAs in melanoma patients, and among them, 17 pairs of co-expressed lncRNAs that could divide SKCM cohort into high-risk and low-risk groups [ 218 ].…”

“…Moreover, the high-risk group is correlated with the expression of specific mutant genes such as BRAF and KIT ( p -value < 0.01) [ 218 ]. Wang et al have used survival analysis and Cox regression model to identify 8 immune-related lncRNAs with prognostic value in SKCM dataset [ 219 ]. Again, MIR205HG expression is linked to the poor outcome ( p -value < 0.001), and HLA-DQB1-AS1 to immune protection ( p -value = 0.048).…”

Construction of a Novel Risk Model Based on the Random Forest Algorithm to Distinguish Pancreatic Cancers with Different Prognoses and Immune Microenvironment Features

“…Immune-related lncRNAs (irlncRNAs) have recently been studied in several cancers. Wang et al [9] identi ed 4 irlncRNAs to establish a risk model for lung adenocarcinoma, while 8 irlncRNAs were used to construct a prognostic model for melanoma [10]. Moreover, irlncRNAs have also utilized in glioblastoma, head and neck squamous cell carcinoma and bladder cancer [11][12][13].…”