Abstract
A computational modeling approach that combined several molecular features yielded a robust breast cancer prognostic model that was independently validated in a new patient data set.