Construction of Prognostic ceRNA Network Landscape in Breast Cancer to Explore Impacting Genes on Drug Response by Integrative Bioinformatics Analysis

Abstract

Background: Breast cancer accounts for 30% of all new female cancers yearly. Bioinformatics serves us to find new biomarkers and facilitate future experimental research. Exploring a distinct network of competing endogenous RNA (ceRNA) that includes potential prognostic, diagnostic, and therapeutic biomarkers is captivating. Methods: Differentially expressed lncRNAs, mRNAs, and miRNAs were collected using Gene Expression Omnibus datasets. DEGs were validated based on TCGA. Functional analysis and pathway activity were also done. Drug sensitivity analyses were done, and IC50 vs. gene expression plots were depicted. Results: A total of 696 mRNAs, 48 lncRNAs, and, 43 miRNAs were identified to have significant differential expression in cancerous breast tissue than normal breast tissue samples. Functional analysis showed significant pathway enrichments in cancer. We found that 13 individual genes, lncRNAs, and miRNAs, CDC6, ERBB2, EZR, HELLS, MAPK13, MCM2, MMP1, SLC7A5, TINCR, TRIP13, hsa-miR-376a, hsa-miR-21, hsa-miR-454 were significantly predictive of poor overall survival and AKAP12, CXCL12, FGF2, IRS2, LINC00342, LINC01140, MEG3, MIR250HG, NAV3, NDRG2, NEAT1, TGFBR3 and, hsa-miR-29c were associated with favorable overall survival. We reached a set of five genes (EGR1, NFIB, TGFBR3, SMARCA4, and MCM2) that exhibit altered expression patterns in breast cancer, resulting in increased susceptibility of cancer cells to certain drug treatments. Conclusion: We successfully made a unique ce-network, providing new clues to understand the regulatory functions of non-coding RNAs (miRNAs and lncRNAs) in the pathogenesis and prognosis of breast cancer and will facilitate further experimental studies to develop new biomarkers in the diagnosis, prognosis and, therapy of breast cancer.