Challenges for Training Translational Researchers in the Era of Ubiquitous Data
The pipeline of stages from T0 to T3 and beyond is attractive and simple, but does not capture the reality that observations at any stage of the translational pipeline may generate hypotheses that are relevant to other stages (including stages that are “earlier”); the observation of a new side effect in a clinical trial may raise the possibility that a new molecular pathway modulates a disease or the observation of a new protein interaction may immediately suggest an unexpected side effect for a new therapeutic. Of course, these crossover observations occur frequently and the scientific enterprise depends on attentive investigators making connections across all parts of the research spectrum. These investigators are essential because they can at once understand the molecular bases of health and disease, how they manifest in human physiological processes, and how measurement and intervention in these processes can alter human disease. The opportunities for integrating data across segments of the research pipeline to both generate and prove hypotheses are myriad, and a few examples include:
In an era of specialization, however, it is not clear how best to create a cadre of investigators who have an understanding of this broad spectrum of biomedical concepts and research specialties that they use to drive biomedical discovery and implementation. Joint MD and PhD training programs create physician scientists with dual perspectives; research experiences for medical students can introduce them to research and prepare them for translational work; and the routine inclusion of “introduction to clinical medicine” within PhD programs can add this perspective to scientific training. Educators have employed these and other interventions to increase the reservoir of scientists who are able to consider broad translational issues—typically while focusing their day‐to‐day work primarily in one segment of the translational research spectrum. Of course, team science also facilitates translational research; scientists bring deep expertise in one segment but develop communication skills to work with experts from other segments to identify important new research opportunities. These communication skills are nontrivial and also require education and nurturing, and perhaps new models for giving credit to translational scientists who build bridges between other specialists.
Associated with this increased activity and focus on translational research is an unprecedented increase in available data from all segments of the translational pipeline. Molecular and genetic data are increasingly freely available for genome, transcriptome, proteome, metabolome, and even large‐scale functional assays. ClinicalTrials.gov not only catalogs the existence of trials but provides useful information about the diseases addressed, compounds testing, diagnostics used, and clinical populations studied. There are numerous efforts to facilitate large‐scale clinical data sharing. Patient advocacy groups are funding the collection and dissemination of data, and there has been a marked rise in the number of very large cohorts of both healthy and diseased patients with large amounts of “real‐world” molecular, sensor, and electronic medical record data. This increase in available data creates unprecedented opportunities to link basic biology to clinical medicine, but also raises substantial challenges in training the next generation of translational research.