Prognostic value of two novel genes SNX10 and PTGDS for cervical squamous cell carcinoma: an integrated bioinformatic analysis

Abstract

KEYWORDSCervical squamous cell carcinoma, methylation prognosis risk model SNX10, PTGDS in females (1). Despite pre-cancerous screening and emerging treatments, CC remains the primary cause of death in women in developing countries (2). When CC becomes metastasizes and recurs, the prognosis gets even worse. Therefore, it is of great significance to establish target treatments of CC based on its to-be-clarified molecular mechanism.Gene expression microarray, as an efficient means of acquiring large-scale genetic data, has been mostly used to study gene expression profiling in many human cancers. New methods and good foreground are provided by microarrays for studying tumor-associated genes, molecular targeting, molecular prediction and therapy. The integration of databases containing gene expression chips allows in-depth study of molecular mechanisms(3, 4).Up to now, the expression profiles of thousands of differentially expressed genes (DEGs) in CC have been researched (5-7). However, the results on some mRNAs are inconsistent. Here we use an unbiased approach to solve this problem.In our study, we screened DEGs from four profiles downloaded from GEO. PPI network was built by STRING Database and hub modules selected via plug-in MCODE. CMap was used to find potential molecules associated with CC. We also validated hub genes with GEO datasets, GEPIA, immunohistochemistry and ONCOMINE. ROC curve analysis and GESA were also done to tease out the significance of hub genes. The flow chart of this research was displayed in Fig. 1. Methods Screening DEGsKeywords "cervical cancer geo accession" were put in the GEO database (https://www.ncbi.nlm.nih.gov/geo/) and the gene expression profiles of GSE6791, GSE63514, GSE39001 and GSE9750 were downloaded. The dataset information is shown in Table 1. We processed unqualified data by R package. The data is calibrated, standardized and log2-transformed.Gene expression analysis was performed using the limma(8) package in the Bioconductor package.Relevant codes were placed into R. We selected four microarray datasets and analyzed them with limma. The |log2fold change (FC)| > 2 and adjusted p < 0.05 were set as cutoffs. RRA package was download (http://cran.r-project.org/)(9) and R was used for running the instruction code.