Applied Research Of Non-coding RNAs Serve As Biomarkers In The Prognosis Prediction Of Gastric Cancer

Gastric cancer is the most common malignant tumor of the digestive system and one of the main causes of cancer-related death.Patients with early gastric cancer have a good prognosis,and the 5-year overall survival rate is about 90%.However,numbers of gastric cancer patients are already in advanced stages when they first diagnosed,and will produce a poor prognosis.Therefore,it is very important to screen prognostic biomarkers and construct signatures to predict prognosis efficiently.ncRNAs is a type of genomic transcript that has no protein-coding ability.A large number of researches shows that ncRNAs plays very critical roles in the development and prognosis of tumors.However,ncRNAs and its signatures have not been systematically studied in the prognostic diagnosis of GC.This study explored the application value of ncRNAs in the prognostic diagnosis of GC.Through bioinformatics analysis,we identified ncRNAs biomarkers related to the prognosis of GC.And further constructed signatures for prognostic evaluation.The whole study was divided into the following two parts:Part ?:Construction of a miRNAs prognostic risk signature in gastric cancerObjective:To identify microRNAs(miRNAs)biomarkers related to prognosis of gastric cancer patients,and construct a miRNAs risk assessment signature for the survival prediction of gastric cancer patients.Methods:miRNAs expression profile data of gastric cancer patients and the relevant clinical information were downloaded from the Cancer Genome Atlas(TCGA)database.Using the "DESeq2" package,differentially expressed miRNAs were identified.Then,univariate Cox regression and Kaplan-Meier analysis were used to identify prognostic miRNAs and multivariate Cox regression was used to construct a prognostic risk assessment model.Throuth "time ROC" package,receiver operating characteristic curve(ROC)curve was used to evaluate the effectiveness of the signature.Then the messenger RNAs(mRNAs)that miRNAs may bind to were predicted by online database,and their possible functions were predicted by gene ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG).Results:Taking |log2Fold Change|>1.and P-Value<0.05 as the standards,248 differentially expressed miRNAs in gastric cancer tissues were identified.Univariate Cox regression and Kaplan-Meier analysis showed that 6 differentially expressed miRNAs had significant correlation with prognosis.Then,the 6 miRNAs were included in the multivariate Cox regression analysis to construct a prognostic risk assessment signature.The risk score=0.04835×miR-181b-1+0.11206×miR-548d-1+0.06800×miR-675+0.07587×miR-708+1.17521×miR-4640+0.08989×m iR-4709.The signature was able to predict the prognostic risk of patients with gastric cancer.The results of kaplan-meier analysis showed that patients with a high risk score had a poor prognosis(P<0.001).The area under the ROC curve(AUC)of the 5-years overall survival rate was 0.776.Conclusion:Through bioinformatics analysis,a miRNAs prognostic risk signature was constructed successfully,and the results of Kaplan-Meier and ROC curve showed that the signature could predict the survival of gastric cancer patients effectively.Part ?:Identification of a 10-lncRNAs signature as potential prognostic biomarker in gastric cancerObjective:To identify the differentially expression lncRNAs in gastric cancer and prognostic lncRNAs biomarkers,and construct a risk assessment model for survival prediction of gastric cancer patients.Methods:Download the gastric cancer RNAs expression data and relevant clinical information corresponding to the samples from the TCGA database.Through the"DESeq2" package to identify differentially expressed RNAs in gastric cancer.Univariate Cox regression analysis and Kaplan-Meier survival analysis were used to identify prognostic lncRNAs.Then the whole data was separated into two part randomly and equally,the training set and the validation set.In the model building phase,the training set data was used.The lncRNAs screened by survival analysis were included in the multivariate Cox regression analysis to construct the prognostic risk assessment model.Use the "timeROC" to draw the ROC curve and evaluate the effectiveness of the model.Then the verification set and the overall data set were used to verify the model.CeRNA network and GO,KEGG analysis were used to predict the underlying mechanism and function of target lncRNAs.Finally,the expression level of LINC01929 was verified by qRT-PCR experiment.In vitro experiments such as real-time cell detection analyzer and migration,invasion were used to verify the biological function of LINC01929.Results:There were a total of 847 differentially expressed lncRNAs,64 differentially expressed miRNAs and 1472 differentially expressed mRNAs(| log2 Fold Change |>2,adjusted P-Value<0.01).And 78 lncRNAs were associated with the overall survival rate of gastric cancer patients.The 10-lncRNAs prognostic risk assessment model was constructed in the training set successfully.And the 3-year and 5-year AUC of the signature were 0.878 and 0.808.The verification results in the validation set and the overall data set confirmed that the signature can predict the survival state of the patient effectively.The function prediction results showed that the 10 lncRNAs could form a ceRNA network through combining with miRNAs and mRNAs to play it's function.The results of GO and KEGG showed that those mRNAs involved in the signaling pathways such as P53,TGF-? and many GO terms such as metabolic process.Knocking down of LINC01929 could inhibit the proliferation,invasion and migration of gastric cancer cells in vitro.Conclusion:Through bioinformatics analysis,lncRNAs prognostic biomarkers of gastric cancer were identified,and constructed a 10-lncRNAs prognostic risk scoring model successfully,which has a good predictive effect on the survival status of patients.