Developing a 5-gene prognostic signature for cervical cancer by integrating mRNA and copy number variations

Abstract

Abstract Background Cervical cancer is frequently detected gynecological cancer all over the world. This study was designed to develop a prognostic signature for an effective prediction of cervical cancer prognosis. Methods Differentially expressed genes (DEGs) were identified based on copy number variation (CNV) data and expression profiles from different databases. A prognostic model was constructed and further optimized by stepwise Akaike information criterion (stepAIC). The model was then evaluated in three groups (training group, test group and validation group). Functional analysis and immune analysis were used to assess the difference between high-risk and low-risk groups. Results The study developed a 5-gene prognostic model that could accurately classify cervical cancer samples into high-risk and low-risk groups with distinctly different prognosis. Low-risk group exhibited more favorable prognosis and higher immune infiltration than high-risk group. Both univariate and multivariate Cox regression analysis showed that the risk score was an independent risk factor for cervical cancer. Conclusions The 5-gene prognostic signature could serve as a predictor for identifying high-risk cervical cancer patients, and provided potential direction for studying the mechanism or drug targets of cervical cancer. The integrated analysis of CNV and mRNA expanded a new perspective for exploring prognostic signatures in cervical cancer.

Keywords

• Cervical cancer
• Copy number variations
• Differential expressed genes
• Prognostic signature
• Bioinformatics