Identification of hub genes in cervical cancer using weighted gene co-expression network analysis

Abstract

Background Cervical cancer ranks second among malignancies in females around the world. Due to the elevated incidence and mortality of this malignancy, deciphering its pathogenesis and identifying related biomarkers is urgently required. Methods First, raw cervical squamous cell carcinoma (CESC) data in GSE63514 were downloaded from the Gene Expression Omnibus (GEO) database. Then, weighted Correlation Network Analysis (WGCNA) was performed to build a co-expression network. Next, comprehensive bioinformatics was performed to determine hub genes, and assess the associated functional annotation, prognostic signature, tumor immunity, DNA mismatch repair, methylation mechanism, candidate molecular drugs, and gene mutations. Results From the key module, ALOX12B, KRT78, RHOD and ZNF750 were selected for validation. K-M plots indicated that these genes had good diagnostic and prognostic values in CESC. Moreover, mutations in these hub genes resulted in the downregulation of most immune genes in CESC. On the other hand, most of the four core genes were negatively correlated with DNA mismatch genes. In addition, we found that RHOD and ZNF750 had decreased methylation in the disease state, while ALOX12B and KRT78 showed no significant differences. Meanwhile, GSVA revealed that most core genes had associations with P53 signaling and the hypoxia signaling pathway. Conclusion WGCNA could identify groups of genes significantly associated with cervical cancer prognosis. These findings provide new insights into CESC pathogenesis, and identify ALOX12B, KRT78, RHOD and ZNF750 as candidate biomarkers for CESC diagnosis and prognosis.