Reactive Oxygen Species (ROS) related signature acted as the immunotherapy clinical response index in melanoma

Abstract

Background: At present, melanoma has become an important factor that seriously threatens people's life, property and health. Oxidative stress is currently believed to affect the prognosis of melanoma patients by affecting the progression of melanoma cells. Method: In this study, the TCGA-SKCM dataset was used to obtain the HTSeq-FPKM RNA-seq transcriptome and clinical data of 471 cutaneous melanoma patients. Subsequently, the GEO database was used to obtain the GSE65904 dataset and GSE120575 single-cell sequencing data, and EMBL-EBI database was used to obtain the transcriptome data of PRJEB23709 as the validation group. In the analysis of single-cell sequencing data, logarithmic normalization was performed on the combined data, and the FindVariableFeatures function was used to identify the first 2000 highly variable genes. All genes were then normalized using the ScaleData function and the dimensions of the data were reduced to 50 principal components by the RunPCA function. Cluster analysis using the "FindNeighbors" and "FindClusters" functions to identify cell clusters at a resolution of 0.1. Next, reduce the dimensions further by selecting the first 50 principal components and applying the UMAP method. CIBERSORT analysis was used to estimate changes in immune cell subpopulations in different groups, and Spearman correlation analysis was used to assess the association between risk scores and immune infiltrating cells. Predictors were selected using LASSO analysis, and prognostic models were constructed by Cox regression analysis. The TIDE approach was used to evaluate the effectiveness of immunotherapy in melanoma patients, and statistical methods were used to analyze the data. Result: In the study, immune cells from melanoma patients were analyzed using the GSE120575 single-cell RNA sequencing dataset and genes associated with ROS were identified. Further studies found that most cell types in the non-responding group had higher ROS marker scores than those in the responding group, and multiple up-regulated gene pathways were present in cells with high ROS markers. By Lasso-Cox regression analysis, a prognostic model based on five ROS-related genes was constructed, and the reliability and validity of the model in the TCGA-SKCM and GSE65904 datasets were verified. In addition, the study found a negative correlation between risk scores and immune-related genes and immune-infiltrating cells, and that patients in the low-risk group responded better to immunotherapy.