Background and purpose: Breast cancer threatens the health of women all over the world. Although a large number of microRNAs (miRNAs) have been found to be abnormally expressed in breast cancer, a complete miRNA-messenger RNA (mRNA) network still needs to be constructed. Methods: The Cancer Genome Atlas (TCGA) database was used to download breast cancer related data sets and analyze the differentially expressed miRNAs between tumor tissues and normal tissues. The miRDB, miRTarBase and StarBase databases were used to analyze the genes targeted by different miRNAs. The ClusterProfiler package in R language was used to enrich and analyze the target genes. String database and Cytoscape 3.6.2 software were used to analyze protein- protein interaction (PPI) network and screen Hub gene. The miRNA-Hub mRNA regulatory network was constructed to determine the research signal axis, and then verified by cell experiments. Results: Two differential miRNAs were identified in TCGA data set; 278 target genes were predicted from the three databases. Ten Hub genes were identified. The constructed miRNA Hub gene network showed that hsa-mir-98-5p/DKK3 axis might play a key role in the progression of breast cancer. Cell functional experiments confirmed that hsa-miR-98-5p could inhibit apoptosis and promote cell proliferation, migration and invasion. The binding of hsa- miR-98-5p to DKK3 was further confirmed by dual luciferase activity assay. Conclusion: In this study, we analyzed a miRNA- mRNA network associated with breast cancer progression and identified an important miRNA mRNA axis in breast cancer.