| Background: Tumor is a mass formed by uncontrolled abnormal proliferation of tissues and cells,which can be divided into benign tumor and malignant tumor.Malignant tumor originating from epithelial tissue is classified as cancer,and that from mesenchymal tissue as sarcoma.Malignant tumor is one of the main factors endangering human health.In 2018,the total number of people dying of malignant tumors all over the world was nearly 10 million.Tumor markers are produced by tumor cells or normal cells affected by tumor cells,which have the properties of representing tumor occurrence and prognosis.Materials and Methods: The tumor RNA-seq data for screening independent prognostic genes and the corresponding clinical data are downloaded from GDC website of TCGA database.For tumors without clinical normal samples,we supplement the normal samples from GTEx database.The general methods are as follows: 1.We cleaned the RNA-seq data and converted the ensemble symbols to gene names;2.We got the differentially expressed genes in tumor samples and normal samples by differential analysis;3.We combined the expression data of differentially expressed genes with survival data to get the differentially expressed genes related to survival;4.After the combination of expression data and clinical data from the first two steps,multivariate Cox analysis was carried out to screen genes that can be used as independent prognostic factors;5.For genes used as independent prognostic factors,we drew the receiver operating characteristic curve(ROC curve)of 5-year survival rate,calculated AUC under the curve,and screened genes with AUC > 0.6 as the final result;6.Analysis of the correlation between independent prognostic genes and clinical symptoms;7.The GO and KEGG enrichment analysis of independent prognostic genes was carries out;8.The online service database was constructed by C#,HTML,CSS,and the results were displayed on the website.Results: We finally screened the independent prognostic genes from 9 tumors.We screened 43 genes in bladder urothelial carcinoma,124 genes in renal clear cell carcinoma,237 genes in renal papillary cell carcinoma,86 genes in pancreatic cancer,20 genes in colon cancer,1194 genes in liver cancer and lung adenocarcinoma 113 genes,21genes in gastric adenocarcinoma and 19 genes in thyroid cancer were screened.We process GO and KEGG enrichment analysis on screened genes,and part of independent prognostic genes were not enriched in GO and KEGG analysis.GO and KEGG enrichment results of tumors have important biological processes such as glycolysis of glucose-6-phosphate and biosynthesis of amino acids.Finally,the differential expression map,survival analysis chart and ROC curve of the genes with independent prognostic function were displayed on the website.Conclusion: Through a series of bioinformatics analysis,we finally obtained a great number of genes with independent prognostic function,including genes expressing protein and genes without corresponding protein,after the screening process.Although some of these genes have been reported to be associated with the corresponding tumors,or have been reported as independent prognostic genes.,most of these genes have not been reported yet.We hope our work can provide help to other researchers.At present,our working data source is entirely from TCGA,and we will add data from other source in the future. |
PDF Full Text Request