Three Survival-Related Genes of Esophageal Squamous Cell Carcinoma Identified by Weighted Gene Coexpression Network Analysis. (9th August 2021)

• Record Type: Journal Article
• Authors: Lu, Di; Wang, He; Wu, Xuanzhen; Zhai, Jianxue; Liu, Xiguang; Dong, Xiaoying; Feng, Siyang; Shi, Xiaoshun; Jiang, Jianjun; Wang, Zhizhi; Chen, Zhiming; Zhao, Shuhua; Zhong, Jinhua; Xiong, Gang; Wu, Hua; Wang, Haofei; Cai, Kaican
• Other Names: Huang Chao Academic Editor.
• Is Part Of: Complexity. Volume 2021(2021)
• Publication Date: 2021-08-09
• Subjects: Chaotic behavior in systems -- Periodicals; Complexity (Philosophy) -- Periodicals; 003
• DOI: 10.1155/2021/9997783
• Languages: English
• ISSNs: 1076-2787

Abstract

Abstract : Background . The aim of this study was to identify novel biomarkers associated with esophageal squamous cell carcinoma (ESCC) prognosis. Methods . 81 ESCC samples collected from The Cancer Genome Atlas (TCGA) were used as the training set, and 179 ESCC samples collected from the Gene Expression Omnibus database (GEO) were used as the validation set. The protein-coding genes of 25 samples from patients who completed the follow-up in TCGA were analyzed to construct a coexpression network by weighted gene coexpression network analysis (WGCNA). Gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analyses were performed for the selected genes. The least absolute shrinkage and selection operator (LASSO) Cox regression model was constructed to analyze survival-related genes, and an optimal prognostic model was developed as well as evaluated by Kaplan–Meier and ROC curves. Results . In this study, a module containing 43 protein-coding genes and strongly related to overall survival (OS) was identified through WGCNA. These genes were significantly enriched in retina homeostasis, antimicrobial humoral response, and epithelial cell differentiation. Besides, through the LASSO regression model, 3 genes (PDLIM2, DNASE1L3, and KRT81) significantly related to ESCC survival were screened and an optimal prognostic 3-gene risk prediction model was constructed. ESCC patients with low and high OS in both sets could be successfully discriminated by calculating a risk score with the linear combination of the expression level of each gene multiplied by the LASSO coefficient. Conclusions . Our study identified three novel biomarkers that have potential in the prognosis prediction of ESCC.