Screening Of Key Genes In Digestive System Malignancies And Construction Of Prognostic CeRNA Regulatory Networks

Malignant tumors of the digestive system(including colorectal carcinoma,gastric cancer,hepatocellular carcinoma,etc.)have high morbidity and mortality in China and even in the world.The occurrence and development mechanism of cancers is not clear enough is one of the important reasons for affecting the treatment effect.Based on the key role of genes,especially for non-codingRNA(ncRNA),in the occurrence and development of malignant tumors,the integrative analysis of multiple transcriptome data will contribute to exploring potential prognostic biomarkers that can provide prognosis or predictive information for patients.This study aims to provide data reference for subsequent cancer treatment by integratively analyzing diverseRNA expression data of the digestive system tumors based on the public databases,and constructing and performing in-depth analysis ceRNA network to understand the prognostic roles according to interaction between mRNA and ncRNA.Firstly,based on transcriptome sequencing data resources from different public databases,multiple data sets were integratively analyzed to obtain differentially expressed gene profiles.Secondly,the protein interaction networks were constructed to screen cancer-related genes combined with survival analysis results,and related miRNA and lncRNA were simultaneously analyzed and screened.Finally,the biological relationship,expression and functional correlation of mRNA,miRNA and lncRNA were performed an integrated analysis,and the regulatory network of ceRNA was constructed and further analyzed to explore the potential prognostic markers in different malignant tumors of the digestive system.Simultaneously,a risk scoring model was constructed and a prognostic analysis system was developed based on the screened key genes in hepatocellular carcinoma.In colorectal carcinoma,a total of 9 key genes were screened,and then 13 miRNAs and 29 lncRNAs were screened,and all of these ncRNAs were differentially expressed with significant prognosis.In-depth analysis showed that AC005154.6:miR-200-b-3p:MFAP5 axis could be used as a prognostic marker of colorectal carcinoma.Of these,AC005154.6 and MFAP5 were significantly down-regulated,indicating that AC005154.6 may bind to miR-200b-3p by competing with MFAP5.The disturbed ceRNA network may contribute to the occurrence and development of cancer.In gastric cancer,a total of 7 key genes were screened,and then 8 associated miRNAs and 13 lncRNAs that were differentially expressed and had significant prognosis were obtained.Further analysis showed that ARHGAP11B:miR-145-5p:GTSE1 axis was a prognostic marker.Based on the integration of multiple data sets by Robust rank aggregation algorithm,a total of 6 key genes were screened in hepatocellular carcinoma,and then 7 related miRNAs and 22 lncRNAs that were differentially expressed with potential prognostic values were screened.After integrated analysis,it was found that SNHG3:miR-139-5p:CENPM axis could be used as a prognostic marker.In addition,a prognostic analysis platform was constructed based on the six screened key genes to evaluate the prognosis of patients according to gene expression.In conclusion,this study discusses the potential prognostic markers in the malignant tumors of digestive system based on an integrative analysis of diverseRNA,and proposes a method to construct ceRNA network using multiple data.We screen and construct ceRNA regulatory networks in the three kinds of malignant tumors of digestive system,and develop a prognostic analysis platform.These results expand the understanding of ceRNA network and the potential roles of ceRNA network as a prognostic indicator in tumorigenesis,which may provide reference data and theoretical support for tracking the disease process and timely adjusting drug treatment strategy.