Identification and validation of a prognostic model for melanoma patients with 9 ferroptosis-related gene signature

Abstract

BackgroundMelanoma is a highly heterogeneous andaggressive cutaneous malignancy. Ferroptosis, a new pathway of cell deathdepending on the intracellar iron, has been shown to be significantlyassociated with apoptosis of a number of tumors, including melanoma.Nevertheless, the relationship between ferroptosis-related genes (FRGs) and themelanoma patients’ prognosis needs to be explored.MethodsDownload expression profiles of FRGs andclinical data from The Cancer Genome Atlas (TCGA) database. 70% data wererandomly selected from the TCGA database and utilized the univariate Coxanalysis and the least absolute shrinkage and selection operator (LASSO)regression model to create a prognostic model, and the remaining 30% was usedto validate the predictive power of the model. In addition, GSE65904 andGSE22153 date sets as the verification cohort to testify the predictive abilityof the signature.ResultsWe identified nine FRGs relating with melanomapatients’ overall survival (OS) and established a prognostic model based ontheir expression. During the research, patients were divided into group ofhigh-risk and low-risk according to the results of LASSO regression analysis.Survival time was significantly longer in the low-risk group than that of in thehigh-risk group (P < 0.001). Enrichment analysis of different risk groupsdemonstrated that the reasons for the difference were related to immune-relatedpathways, and the degree of immune cell infiltration in the low-risk group wassignificantly higher than that in the high-risk group.ConclusionsThe FRG prognostic model we established canpredict the prognosis of melanoma patients and may further guide subsequenttreatment.