At a Glance:
PhD in theoretical computer science from the …cole Normale SupÈrieure in Paris.
Prior to joining Hybrigenics served at the French Embassy in Washington, DC, tracking the software and telecommunications industries.
Co-founder of the Bio-Pathways Consortium.
Lives in Paris. Enjoys movies, Kung Fu, and traveling. Avid reader of philosophy and sociology.
QWhere will bioinformatics be in two years? Five years?
AWhat were seeing in bioinformatics right now is an effort to process and manage data coming from a number of different technologies. Were also seeing the need to integrate these different data types.
The first thrust, of managing and processing, and maybe normalizing data from the different platforms, will continue to evolve over the next two years, fueling the second effort, data integration. There are two levels to this integration effort. One level is more technological: acquiring these data types, storing them in a way that allows you to query and visualize them together. This is one area that will see a lot of progress over the next two years, with the emergence of accepted practices and some standards.
Five years from now we can be a bit bolder. The scientific part of bioinformatics is going to become increasingly important, driving the second level of integration. That means going beyond just looking at one data type or looking at even several data types together. Its looking at this information structured by certain models and using these to make a biological hypothesis, test the hypothesis, and use the results of predictions to plan experiments and go back to the bench in an iterative process.
QWhat are the biggest challenges the bioinformatics sector faces?
AThe immediate challenge is integrating the data at the IT level standards, normalizing the data, and otherwise converging on good practices for using someone elses data in a way that can generate confidence in your results.
The second challenge is a much more difficult one to move from the local gene- or protein-centric view of bioinformatics that is looking at data linked to just one object to looking at more global models of pathways or large parts of cellular mechanisms.
Another challenge will be the adoption of bioinformatics into mainstream biology. Bioinformatics will have to be adopted not only as a set of tools that a biologist can use but also as a valuable scientific approach to biological problems.
A final challenge is training bioinformaticians. We dont have enough people with dual backgrounds. On the other hand, we dont want to train blindly for todays needs, which are not necessarily adequate for the future. Also, providing informatics support for biologists, developing industrial-grade software, or conducting original research in computational biology may require different profiles.
QWith what companies do you have partnerships?
AWe have partnerships with biotech companies Oxford GlycoSciences, Lynx, XTL Biopharmaceuticals, and Genome Express; with a pharmaceutical company, Servier; and we also have very strong ties to academia in France: with the Pasteur Institute, with the National Center of Computer Science, and with the Curie Institute. The majority of these partnerships are focused on particular therapeutic areas.
QDo you expect to see more M&A activity in the sector?
AYes. There is pressure to move downstream in the drug discovery process for biotech companies as well as bioinformatics companies. Companies that focused primarily on a certain type of software linked to a certain type of data may need to align themselves or get acquired by other companies in order to reach critical mass.
QWhat made you decide to enter a career in bioinformatics?
AMy background was in math and theoretical computer science, with some cognitive science. When I had met the biologists who were starting Hybrigenics I realized I had found the perfect field in bioinformatics. You can apply fairly advanced theory to interesting and complex real-world systems, yet applications are very concrete and thats quite attractive.