The increasing number of clinical and biological parameters that need to be explored to achieve precision medicine makes it almost impossible to design dedicated trials. 1 New approaches are needed for all populations of patients. By 2020, a medical decision will rely on up to 10,000 parameters for a single patient, 2 but it is traditionally thought that our cognitive capacity can integrate only up to five factors in order to make a choice. Clinicians will need to combine clinical data, medical imaging, biology, and genomics to 2achieve state-of-the-art radiotherapy. Although sequencing costs have significantly decreased, 3,4 we have seen the generalization of electronic health records (EHRs) and record-and-verify systems that generate a large amount of data. 5 Data science has an obvious role in the generation of models that could be created from large databases to predict outcome and guide treatments. A new paradigm of data-driven decision making: The reuse of routine health care data to provide decision support is emerging. To quote I. Kohane, “Clinical decision support algorithms will be derived entirely from data … The huge amount of data available will make it possible to draw inferences from observations that will not be encumbered by unknown confounding.” 6