Fig. 2

The shifting paradigm of translational human models. In the past, animal models were almost exclusively used for pre-clinical analyses, with limited success in translation to humans. NHP often served as a more relevant model for safety testing prior to attempts to test in humans, although on rare occasions this led to unanticipated and devastating effects in human trials. Currently, more strategies are incorporated into translational models, including sampling from people for in vitro assays. The data derived from human ex vivo and in vitro testing is often used to inform animal models and vice versa. As more high throughput data are made publicly available, computational models can contribute to the translational effort as well. In the future, it may be possible to bypass animal models entirely as more information is gathered from a variety of people of diverse health, genetic, and environmental backgrounds. As we gather broad data from human cohorts, our hope is that our predictive abilities and computational models will improve such that we no longer rely on animal models, although they will undoubtedly continue to play at least a supplemental role in translation