| Objectives Based on RNA sequencing data and clinical feature data of gastric cancer and adjacent tissues in the network database,bioinformatics analysis was performed using system biology methods combined with machine learning algorithms to screen out the long noncoding RNA(lncRNA)molecules of gastric cancer prognosis markers,which provides a reference for studying the molecular mechanism of gastric cancer development.Methods 1 Downloading the RNA-sequence data and clinical characteristics of gastric cancer and paracancerous tissues from the official website of the Cancer Genome Atlas(TCGA).It was need to extract lncRNA data and perform standardization processing in the data.The “edge R” function package was used to screen differentially expressed lncRNAs of gastric cancer.The Weighted Gene Correlation Network Analysis(WGCNA)identified lncRNA modules related to the overall survival time of gastric cancer.2 By the data of lncRNA in the module,it was the way to construct gastric cancer prognosis lncRNA model that using the Least Absolute Shrinkage and Selection Operator(LASSO).Drawing the Receiver Operating Characteristic curve(ROC),and calculating the area under the curve(AUC)were used to evaluate model.According to the model,the risk scores corresponding to gastric cancer and paracancerous samples were calculated.Gastric cancer was divided into high-risk group and low-risk group in the median risk score.Kaplan-Meier method was used for survival analysis to evaluate the model’s ability to predict the prognosis of gastric cancer.It was helpful to identify the key lncRNA molecules for gastric cancer prognosis in a single-gene ROC curve,AUC and survival analysis.Starbase database and Mutil Experiment Matrix(MEM)database were used to predict key lncRNA target genes for gastric cancer prognosis.Results 1 In this study,345 cases of gastric cancer and 32 cases of paracancerous tissue are included in RNA sequencing data and clinical information data.3301 gastric cancer differentially expressed lncRNAs are screened out,of which 2439 lncRNA expressions were up-regulated and 862 lncRNA expressions are down-regulated.In the help of WGCNA algorithm,the differentially expressed lncRNAs are applied to construct a weighted gene co-expression network,which is divided into 17 modules.The green module is significantly related to the overall survival time of gastric cancer.2 Using the LASSO algorithm to extract features from 100 lncRNAs in the green module,and screen out 11 lncRNAs to construct a lncRNA prognosis model of gastric cancer.The survival analysis results of the model show that the 5-year overall survival time of the high-risk group is shorter than that of the low-risk group.The AUC of the model is 0.641,which indicates that the model has good prediction performance.Calculating the AUC of a single lncRNA in the model,which shows that the maximum AUC of LINC00665 is 0.695,suggesting that LINC00665 can effectively predict the prognosis of gastric cancer independently.Survival analysis reveals differences in survival time between gastric cancer high and low expression groups divided by LINC00665(P=0.007).LINC00665 may be one of prognostic molecular markers of gastric cancer.Using star Base database and MEM database for target gene prediction analysis,it is found that HKR1 may be the target gene of LINC00665.Correlation analysis finds that LINC00665 and HKR1 are closely related in gastric cancer(r = 0.401,P = 6.93e-16).Conclusions 1 Based on the TCGA and WGCNA algorithm,the lncRNA module of gastric cancer prognosis containing 100 lncRNA is identified,which provides data support for screening lncRNA molecular markers related to gastric cancer prognosis.2 Based on the LASSO algorithm,it is screened that LINC00665 is a possible lncRNA molecular marker for predicting the prognosis of gastric cancer and may affect the occurrence and development of gastric cancer by regulating the target gene HKR1.Figure 18;Table 4;Reference 135... |
PDF Full Text Request