The estrogen receptor (ER)-negative breast cancer subtype is aggressive with few treat-ment options available. To identify specific prognostic factors for ER-negative breast cancer, this study included 705,729 and 1034 breast invasive cancer patients from the Surveillance, Epidemiol-ogy, and End Results (SEER) and The Cancer Genome Atlas (TCGA) databases, respectively. To identify key differential kinase-substrate node and edge biomarkers between ER-negative and ER-positive breast cancer patients, we adopted a network-based method using correlation coefficients between molecular pairs in the kinase regulatory network. Integrated analysis of the clinical and molecular data revealed the significant prognostic power of kinase-substrate node and edge featuresfor both subtypes of breast cancer. Two promising kinase-substrate edge features, CSNK1A1-NFATC3 and SRC-OCLN, were identified for more accurate prognostic prediction in ER-negative breast cancer patients.