Identification of key molecular targets that correlate with breast cancer through bioinformatic methods

The present study aimed to identify key molecular targets of breast cancer for targeted treatment and to improve the survival rate.

Overlapped difference expression genes in three datasets were identified in a weighted gene co-expression network analysis (WGCNA) module and MetaDE.ES analysis. Combined with the prognosis information [time, death, status and relative survival (RS)] in GSE42568, single-factor Cox regression analysis was used to screen the genes that were significantly related to the prognosis in the target gene set.

In total, 13 optimal gene combinations with a significantly correlated prognosis were obtained, including SSPN, NELL2, AGTR1, NRIP3, IKZF2, NAT1, CXCL12, NPY1R, PRAME, PPP1R1B, CRISP3, NMU and GSTP1. In addition, there was a significant correlation between the samples given by the prognostic prediction system and the validation dataset (GSE20685 and TCGA), with p values of 0.0299 in GSE20685 and 1.461 × 10^-5 in TCGA, and an area under the receiver operating characteristic of 0.942 and 0.923, respectively. RS-related differentially expressed genes between high- and low-risk groups were significantly related to biological processes such as cell period and the hormone stimulation response, and were also significantly involved in KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways such as cell period, the peroxisome proliferator-activated receptor signaling pathway and the cancer pathway.

By predicting the survival risk of breast cancer patients based on the 13 optimal genes, high-risk patients would be detected early. Accordingly, this would help in the formulation of an appropriate treatment plan for patients.