Identification of novel prognostic genes of triple-negative breast cancer using meta-analysis and weighted gene co-expressed network analysis

BACKGROUND: Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer with high rates of metastasis and recurrence. Conventional clinical treatments are ineffective for it as it lacks therapeutic biomarkers. Figuring out the biomarkers related to TNBC will be beneficial for its clinical treatment and prognosis. METHODS: Five independent datasets downloaded from the Gene Expression Omnibus database were merged to identify differentially expressed genes between TNBC and non-TNBC samples by using the MetaDE.ES method followed by mapping the differentially expressed genes into a protein-protein interaction network. Meanwhile, the weighted gene co-expressed network analysis (WGCNA) of The Cancer Genome Atlas data was performed to screen the hub genes. The gene functional analyses were conducted by Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. The correlation between gene expression level and patient overall survival was evaluated by survival analysis. RESULTS: A total of 11 differentially expressed genes (CDH1, SP1, MYC, FAF2, IFI16, MDM2, AR, DBN1, HSPB1, FLNA, YWHAB) were obtained from the protein-protein interaction network with degree >10. WGCNA revealed 5 hub genes (TPX2, CTPS1, KIF2C, MELK, CDCA8) that were significantly associated with TNBC. Cell cycle, oocyte meiosis, spliceosome were the pathways significantly enriched in these genes according to GO functionally annotated terms and KEGG pathways analysis. The Kaplan-Meier curves showed that the expression levels of HSPB1, IFI16, TPX2 were significantly associated with the survival time of TNBC patients (P<0.05). CONCLUSIONS: A total of 16 genes significantly associated with TNBC were identified by bioinformatic analyses. Among these 16 genes, HSPB1, IFI16, TPX2 might be able to be used as biomarkers of TNBC.