Current biological discourse is replete with discussions over the significance of distinguishing between 'data-driven' and 'hypotheses-driven' research. This paper recognizes that there is a difference between these two modes of research, but questions the idea that theory plays no significant role in data-driven research. I focus on a key tool for the classification and distribution of data across research contexts: the 'controlled vocabularies' referred to as bio-ontologies. Bio-ontologies constitute a representation of the knowledge that makes it possible for researchers to use data as evidence within specific research contexts. In this sense, they are crucial theoretical components of data-driven research.
Sabina Leonelli is a research fellow of the ESRC Centre for Genomics in Society (Egenis) based at the University of Exeter and a visiting fellow in the Leverhulme/ESRC project 'How Well Do 'Facts' Travel?' based at the London School of Economics. She was trained in the history, philosophy, and social studies of science in London and Amsterdam. Her current work focuses on the relations between regulatory and classificatory practices within the biomedical sciences, with particular attention to the role played by bioinformatic tools for data sharing. She is also writing a monograph on the history of research on the model organism Arabidopsis thaliana.