A Model-Based Prognostic Predictor for Urothelial Bladder Carcinoma Through the Integrated Bioinformatics Analysis

The aim of this study is to identify potential biomarkers for early diagnosis of bladder cancer to improve survival. Bladder cancer is a common cancer disease worldwide, which a large proportion of types is urothelial cells. The purpose of our research is to use bioinformatics to find out the feature genes that cause urothelial carcinoma. In our study, we download the BLCA gene expression profile from the cancer genome atlas (TCGA). According to a screening of differentially expressed analysis, the 5035 Differentially Expressed Genes were selected (DEGs) associating with the urothelial carcinoma to construct gene co-expression networks. With weighted correlation network analysis (WGCNA), the study selected the most key module which had a similar expression for analysis. After that, we obtained hub genes that were highly related to Urothelial Carcinoma (UC) and identified the significant enrichment pathways of genes. Based on Least Absolute Selection and Shrinkage Operator (LASSO) regression model, hub genes that had less influence decaying to zero. Lastly, two feature genes (STIL, GINS1) were screened. To validate the clinical outcomes of the feature genes, different types of data sets were used to analyze the overall survival of the genes. Gene STIL was significantly related to patient prognosis. This research will help in the development of Urothelial Carcinoma, this gene can be directed to explore new treatment strategies or innovative diagnostic methods.