Breast cancer is the most common non-cutaneous cancer diagnosed in women in the United States and the second most common cause of cancer-related mortality. Over the past two decades, the progress in screening and adjuvant systemic therapies noticeably improved the survival rate. However, traditional methods of characterizing tumors are imprecise and create heterogeneous groupings of tumors and patients. As a result, despite the important medical advances in diagnosis and treatment of breast cancer, one-third of the patients with initial breast tumor have recurrence of the disease 10 years after the diagnosis. Therefore, novel tools for discovery of strong prognostic and predictive markers that can be used to identify patients at high risk for relapse and aid in the selection of the most appropriate therapy are needed. This review analyzes some recent achievements in the development of such tools.