Dataset Information

Cuproptosis-related risk score based on machine learning algorithm predicts prognosis and characterizes tumor microenvironment in head and neck squamous carcinomas.

ABSTRACT: Cuproptosis is a recently discovered type of programmed cell death that shows significant potential in the diagnosis and treatment of cancer. It has important significance in the prognosis of HSNC. This study aims to construct a cuproptosis-related prognostic model and risk score through new data analysis methods such as machine learning algorithms for the prognosis analysis of HSNC. Protein-protein interaction network and machine learning methods were employed to identify hub genes that were used to construct a TreeGradientBoosting model for predicting overall survival. The relationship between the risk scores obtained from the model and features such as tumor microenvironment (TME) and tumor immunity was explored. The C-indexes of the TreeGradientBoosting model in the training and validation cohorts were 0.776 and 0.848, respectively. The nomogram based on risk scores and clinical features showed good performance, and distinguished the TME and immunity between high-risk and low-risk groups. The cuproptosis-associated risk score can be used to predict prognoses, TME, and tumor immunity of HNSC patients.

SUBMITTER: Ye M

PROVIDER: S-EPMC10363129 | biostudies-literature | 2023 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Cuproptosis-related risk score based on machine learning algorithm predicts prognosis and characterizes tumor microenvironment in head and neck squamous carcinomas.

Ye Maodong M Zhang Guangping G Lu Yongjian Y Ren Shuai S Ji Yingchang Y

Scientific reports 20230722 1

Cuproptosis is a recently discovered type of programmed cell death that shows significant potential in the diagnosis and treatment of cancer. It has important significance in the prognosis of HSNC. This study aims to construct a cuproptosis-related prognostic model and risk score through new data analysis methods such as machine learning algorithms for the prognosis analysis of HSNC. Protein-protein interaction network and machine learning methods were employed to identify hub genes that were us ...[more]

PMID: 37481622

Similar Datasets

Project description:IntroductionCuproptosis is a novel copper-dependent regulatory cell death (RCD), which is closely related to the occurrence and development of multiple cancers. However, the potential role of cuproptosis-related genes (CRGs) in the tumor microenvironment (TME) of colon adenocarcinoma (COAD) remains unclear.MethodsTranscriptome, somatic mutation, somatic copy number alteration and the corresponding clinicopathological data of COAD were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus database (GEO). Difference, survival and correlation analyses were conducted to evaluate the characteristics of CRGs in COAD patients. Consensus unsupervised clustering analysis of CRGs expression profile was used to classify patients into different cuproptosis molecular and gene subtypes. TME characteristics of different molecular subtypes were investigated by using Gene set variation analysis (GSVA) and single sample gene set enrichment analysis (ssGSEA). Next, CRG Risk scoring system was constructed by applying logistic least absolute shrinkage and selection operator (LASSO) cox regression analysis and multivariate cox analysis. Real-time quantitative polymerase chain reaction (RT-qPCR) and immunohistochemistry (IHC) were used to exam the expression of key Risk scoring genes.ResultsOur study indicated that CRGs had relatively common genetic and transcriptional variations in COAD tissues. We identified three cuproptosis molecular subtypes and three gene subtypes based on CRGs expression profile and prognostic differentially expressed genes (DEGs) expression profile, and found that changes in multilayer CRGs were closely related to the clinical characteristics, overall survival (OS), different signaling pathways, and immune cell infiltration of TME. CRG Risk scoring system was constructed according to the expression of 7 key cuproptosis-related risk genes (GLS, NOX1, HOXC6, TNNT1, GLS, HOXC6 and PLA2G12B). RT-qPCR and IHC indicated that the expression of GLS, NOX1, HOXC6, TNNT1 and PLA2G12B were up-regulated in tumor tissues, compared with those in normal tissues, and all of GLS, HOXC6, NOX1 and PLA2G12B were closely related with patient survival. In addition, high CRG risk scores were significantly associated with high microsatellite instability (MSI-H), tumor mutation burden (TMB), cancer stem cell (CSC) indices, stromal and immune scores in TME, drug susceptibility, as well as patient survival. Finally, a highly accurate nomogram was constructed to promote the clinical application of the CRG Risk scoring system.DiscussionOur comprehensive analysis showed that CRGs were greatly associated with TME, clinicopathological characteristics, and prognosis of patient with COAD. These findings may promote our understanding of CRGs in COAD, providing new insights for physicians to predict prognosis and develop more precise and individualized therapy strategies.

Project description:Background: Cuproptosis is a new type of cell death that induces protein toxic stress and eventually leads to cell death. Hence, regulating cuproptosis in tumor cells is a new therapeutic approach. However, studies on cuproptosis-related long noncoding RNA (lncRNA) in head and neck squamous cell carcinoma (HNSC) have not been found. This study aimed to explore the cuproptosis-related lncRNAs prognostic marker and their relationship to immune microenvironment in HNSC by using bioinformatics methods. Methods: RNA sequencing, genomic mutations, and clinical data of TCGA_HNSC were downloaded from The Cancer Genome Atlas. HNSC patients were randomly assigned to either a training group or a validation cohort. The least absolute shrinkage and selection operator Cox regression and multivariate Cox regression models were used to determine the prognostic model in the training cohort, and its independent prognostic effect was further confirmed in the validation and entire cohorts. Results: Based on previous literature, we collected 19 genes associated with cuproptosis. Afterward, 783 cuproptosis-related lncRNAs were obtained through coexpression. Cox model revealed and constructed eight cuproptosis-related lncRNAs prognostic marker (AL132800.1, AC090587.1, AC079160.1, AC011462.4, AL157888.1, GRHL3-AS1, SNHG16, and AC021148.2). Patients were divided into high- and low-risk groups based on the median risk score. The Kaplan-Meier survival curve revealed that the overall survival between the high- and low-risk groups was statistically significant. The receiver operating characteristic curve and principal component analysis demonstrated the accurate prognostic ability of the model. Univariate and multivariate Cox regression analysis showed that risk score was an independent prognostic factor. In addition, we used multivariate Cox regression to establish a nomogram of the predictive power of prognostic markers. The tumor mutation burden showed significant differences between the high- and low-risk groups. HNSC patients in the high-risk group responded better to immunotherapy than those in the low-risk group. We also found that risk scores were significantly associated with drug sensitivity in HNSC. Conclusion: In summary, our study identified eight cuprotosis-related lncRNAs signature of HNSC as the prognostic predictor, which may be promising biomarkers for predicting the benefit of HNSC immunotherapy as well as drug sensitivity.

Dataset Information

Cuproptosis-related risk score based on machine learning algorithm predicts prognosis and characterizes tumor microenvironment in head and neck squamous carcinomas.

Publications

Cuproptosis-related risk score based on machine learning algorithm predicts prognosis and characterizes tumor microenvironment in head and neck squamous carcinomas.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets