A perspective on the selection of experimental models

We are entering an age when advances in biotechnology, greater communication among teams of scientists using multiple model systems, and the power of the computer will dwarf the great advances of the mid-20th century. Advances in medicine rest largely on the use of models of biological systems, including animal models. Model systems used by most biomedical scientists can be considered in four categories: theoretical, in vitro, nonmammalian and mammalian. Each category has advantages and limitations in describing the dynamic milieu of events that characterize human biologic response. Although individual models can be good predictors, multiple models are better than single models; the most critical drawback is lack of human information for comparison. The scientific community must improve the methodologies necessary to extrapolate data from the various model systems to the human, because these model systems are the foundation of discovery in biomedicine and the bedrock upon which to base valid risk assessments. It is vital that we communicate the importance of these techniques to the general public. Advances in science and improved quality of life for the human race depend on these principles.