Screening and Identification of Key Biomarkers of Gastric Cancer: Three Genes Jointly Predict Gastric Cancer.

Meng-Jie Shan,Peng Guo,Dexian Kong,Yuan-Meng Zhang,Ya-Bin Liu,Ling-Bing Meng

doi:10.3389/fonc.2021.591893

Abstract

BackgroundGastric cancer (GC) is one of the most common cancers all over the world, causing high mortality. Gastric cancer screening is one of the effective strategies used to reduce mortality. We expect that good biomarkers can be discovered to diagnose and treat gastric cancer as early as possible.MethodsWe download four gene expression profiling datasets of gastric cancer (GSE118916, GSE54129, GSE103236, GSE112369), which were obtained from the Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) between gastric cancer and adjacent normal tissues were detected to explore biomarkers that may play an important role in gastric cancer. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of overlap genes were conducted by the Metascape online database; the protein-protein interaction (PPI) network was constructed by the STRING online database, and we screened the hub genes of the PPI network using the Cytoscape software. The survival curve analysis was conducted by km-plotter and the stage plots of hub genes were created by the GEPIA online database. PCR, WB, and immunohistochemistry were used to verify the expression of hub genes. A neural network model was established to quantify the predictors of gastric cancer.ResultsThe relative expression level of cadherin-3 (CDH3), lymphoid enhancer-binding factor 1 (LEF1), and matrix metallopeptidase 7 (MMP7) were significantly higher in gastric samples, compared with the normal groups (p<0.05). Receiver operator characteristic (ROC) curves were constructed to determine the effect of the three genes’ expression on gastric cancer, and the AUC was used to determine the degree of confidence: CDH3 (AUC = 0.800, P<0.05, 95% CI =0.857-0.895), LEF1 (AUC=0.620, P<0.05, 95%CI=0.632-0.714), and MMP7 (AUC=0.914, P<0.05, 95%CI=0.714-0.947). The high-risk warning indicator of gastric cancer contained 8<CDH3<15 and 10<expression of LEF1<16.ConclusionsCDH3, LEF1, and MMP7 can be used as candidate biomarkers to construct a neural network model from hub genes, which may be helpful for the early diagnosis of gastric cancer.

Highlights

Gastric cancer (GC) is one of the most common cancers all over the world, causing high mortality
We found that aldo-keto reductase family 1 member C1 (AKR1C1), CDH3, lymphoid enhancer-binding factor 1 (LEF1), SLIT3, matrix metallopeptidase 7 (MMP7), and 15-hydroxyprostaglandin dehydrogenase (HPGD) genes were significantly correlated with prognosis (Figure 9)
Receiver operator characteristic (ROC) curves were constructed to determine the effect of the three genes’ expression on gastric cancer, and the AUC was used to determine the degree of confidence: CDH3 (AUC = 0.800, P

Summary

Introduction

Gastric cancer (GC) is one of the most common cancers all over the world, causing high mortality. Gastric cancer screening is one of the effective strategies used to reduce mortality. The treatment of gastric cancer has limited effect. The survival of some patients with advanced gastric cancer can be prolonged through chemotherapy, most chemotherapy has limited efficacy and a short maintenance time. The diagnosis of the tumor molecular and targeted treatment is an effective means for improving the early diagnosis rate [3]. It is the direction and goal of the development of the clinical treatment of the tumor, and the achievement of this goal will depend on the search for specific tumor biomarkers [4]

Methods

Results

Conclusion