Abstract
PURPOSE: Breast cancer is one of the most common malignancies among females, and its prognosis is affected by a complex network of gene interactions. Weighted gene co-expression network analysis was used to construct free-scale gene co-expression networks and to identify potential biomarkers for breast cancer progression. METHODS: The gene expression profiles of GSE42568 were downloaded from the Gene Expression Omnibus database. RNA-sequencing data and clinical information of breast cancer from TCGA were used for validation. RESULTS: A total of ten modules were established by the average linkage hierarchical clustering. We identified 58 network hub genes in the significant module (R(2) = 0.44) and 6 hub genes (AGO2, CDC20, CDCA5, MCM10, MYBL2, and TTK), which were significantly correlated with prognosis. Receiver-operating characteristic curve validated that the mRNA levels of these six genes exhibited excellent diagnostic efficiency in the test data set of GSE42568. RNA-sequencing data from TCGA showed that the expression levels of these six genes were higher in triple-negative tumors. One-way ANOVA suggested that these six genes were upregulated at more advanced stages. The results of independent sample t test indicated that MCM10 and TTK were associated with tumor size, and that AGO2, CDC20, CDCA5, MCM10, and MYBL2 were overexpressed in lymph-node positive breast cancer. CONCLUSIONS: AGO2, CDC20, CDCA5, MCM10, MYBL2, and TTK were identified as candidate biomarkers for further basic and clinical research on breast cancer based on co-expression analysis.