Biology is generating very large amounts of data. Today biological databases collect data produced by research in biology everywhere in the world. These databases are crucial in many fields of life science research, including medical research (Attwood et al., 2011). The Pharmaceutical industry has faced a crisis of innovation for more than 30 years, EVEN though the investment in R&D activities has not stopped increasing. Now they rely the more and more on the use of databases and their research has become data-driven. There are several reasons to believe that this shift is likely to come with an increase in creativity which would solve the innovation crisis of the industry. Research in scientific creativity highlights the importance of variety and knowledge distance. We will look at whether the use of databases allows pharmaceutical companies to access more distant knowledge. To do so we use qualitative case study methodology and interrogate nine pharmaceutical companies who use the Large European Public databases EMBL-EBI. The interviewees are all managers of bioinformatics departments of their companies. This position gives them a privileged view as they are the intermediaries between the researchers who use the databases daily and the databases themselves. Our findings can be summarized as follows. These databases are a crucial resource as they offer access to a high variety of data. This variety of data comes in the form of variety in its origin (mostly data produced by other communities) and on its typology (kinds of compounds and parameters associated). Additionally to this variety, databases allow scientists to make new connections between knowledge bases. Concerning the kinds of creative output that one kind find, it consists on one side on promising compounds and on the other side on the opening of new research questions and new research paths. Finally, we find an unexpected result which is that databases allow for serendipity, as it exposes a big amount of information to the entire scientific community. We develop all these aspects in the following lines to better answer to the research questions that were asked before.
