Identification of hub genes and drug-gene interactions for targeted breast cancer treatment by integrated bioinformatics analysis

Breast cancer (BC) is one of the most common cancer types in women. In addition to conventional methods for BC diagnosis, applying methods for a fast and accurate prognosis at the early stage of cancer is very meaningful for the treatment of the disease. To date, the most advanced methods are molecular diagnostics and bioinformatics. In this study, bioinformatics is applied to genetic testing for BC diagnosis; namely the R programming language combined with the bioinformatics toolkit was used to analyze gene expression levels between normal and tumor tissues in three gene expression profiles (GSE29431, GSE42568, GSE21422). The bioinformatics approaches included identification of differentially expressed genes (DEGs) and hub genes, Gen Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) analyses, the construction of a protein-protein interaction (PPI) network, and module analysis. Following the completion of the hub gene selection process, coexpression and survival analysis were carried out. Finally, the GEPIA2 and DGIdb databases were utilized to verify the expression levels of hub genes and select the candidate drugs for BC, respectively. A total of 1369 DEGs was identified, including 400 upregulated DEGs and 969 downregulated DEGs. Thereafter, 10 hub genes (CDK1, CCNA2, CCNB1, CCNB2, TOP2A, KIF11, RRM2, BUB1B, CDC20, and NCAPG) were identified as potential biomarkers for BC diagnosis, prognosis, and therapy. Six screened small molecules, dexrazoxane, teniposide, amsacrine, etoposide, mitoxantrone and daunorubicin, were determined to be the new targeted drugs for BC treatment.