Interactive learning: lessons from two hybrids over two decades.

The initial yeast two-hybrid experiment - published in 1989 - described an approach to detecting protein-protein interactions that has flourished over the last two decades, leading to the assembly of large-scale data sets of these interactions. Yet the yeast assay originated because of the laboratory's interests in technology development, not because of its need to identify partners of any protein then under study. In addition to such motivating forces, other features of the process of originating a technology can be revealed by considering the lessons of the two-hybrid approach. These include the value of timeliness in a method's development, the willingness of an investigator to try experimental approaches that prove fruitless, the ability of biological macromolecules to display surprising attributes, the benefits of a community expending efforts to expand the uses of a technology platform, and the role of scientific training of those who work in technology.