Identifying TME signatures for cervical cancer prognosis based on GEO and TCGA databases

Wen-Tao Xia; Wang-Ren Qiu; Wang-Ke Yu; Zhao-Chun Xu; Shou-Hua Zhang

doi:10.1016/j.heliyon.2023.e15096

Identifying TME signatures for cervical cancer prognosis based on GEO and TCGA databases

Heliyon. 2023 Apr 7;9(4):e15096. doi: 10.1016/j.heliyon.2023.e15096. eCollection 2023 Apr.

Authors

Wen-Tao Xia¹, Wang-Ren Qiu¹, Wang-Ke Yu¹, Zhao-Chun Xu¹, Shou-Hua Zhang²

Affiliations

¹ School of Information Engineering, Jingdezhen Ceramic University, Jingdezhen, China.
² Department of General Surgery, Jiangxi Provincial Children's Hospital, Nanchang, China.

Abstract

The mortality rate from cervical cancer (CESC), a malignant tumor that affects women, has increased significantly globally in recent years. The discovery of biomarkers points to a direction for the diagnosis of cervical cancer with the advancement of bioinformatics technology. The goal of this study was to look for potential biomarkers for the diagnosis and prognosis of CESC using the GEO and TCGA databases. Because of the high dimension and small sample size of the omic data, or the use of biomarkers generated from a single omic data, the diagnosis of cervical cancer may be inaccurate and unreliable. The purpose of this study was to search the GEO and TCGA databases for potential biomarkers for the diagnosis and prognosis of CESC. We begin by downloading CESC (GSE30760) DNA methylation data from GEO, then perform differential analysis on the downloaded methylation data and screen out the differential genes. Then, using estimation algorithms, we score immune cells and stromal cells in the tumor microenvironment and perform survival analysis on the gene expression profile data and the most recent clinical data of CESC from TCGA. Then, using the 'limma' package and Venn plot in R language to perform differential analysis of genes and screen out overlapping genes, these overlapping genes were then subjected to GO and KEGG functional enrichment analysis. The differential genes screened by the GEO methylation data and the differential genes screened by the TCGA gene expression data were intersected to screen out the common differential genes. A protein-protein interaction (PPI) network of gene expression data was then created in order to discover important genes. The PPI network's key genes were crossed with previously identified common differential genes to further validate them. The Kaplan-Meier curve was then used to determine the prognostic importance of the key genes. Survival analysis has shown that CD3E and CD80 are important for the identification of cervical cancer and can be considered as potential biomarkers for cervical cancer.

Keywords: Biomarkers; Cervical cancer; DNA methylation; Differentially expressed genes; Tumor microenvironment.