Inria Awards 2015

AB -

Marc Lavielle : Inria – French Académie des Sciences – Dassault Systèmes Innovation Award

© Inria / C. Tourniaire

Using statistics as a modern tool in order to better address the major health issues of our time: therein lies the challenge that Marc Lavielle has set himself. By reconciling high-level scientific work and a unique ability to build on industrial partnerships, he has succeeded in developing an innovative solution that today benefits the entire pharmaceutical industry.

Some careers occasionally take an unexpected turn, and Marc Lavielle is a perfect example of this. Today he is an Inria research director leading the Popix team, and yet he started his professional life as... a typographer. After three years in the printing world, Marc Lavielle eventually decided to change direction, and so in 1981 he went to the Paris Sud University in Orsay. On the agenda was a diploma of higher education and a degree in mathematics, followed by a post-graduate diploma in statistics. He then left for the Central University of Venezuela in Caracas, where he wrote a PhD on deconvolution and rupture detection methods in geophysics. In parallel, he occupied the position of assistant professor in statistics.

A team effort

At the beginning of the 1990s, Marc Lavielle returned to France and became a senior lecturer at Paris Descartes University. Then, in 1998, he became a professor at Paris Descartes IUT (University Institute of Technology). "I have worked on new modelling algorithms since the beginning of my thesis , explains Marc Lavielle.These tools and associated methods have proven to have real potential in the field of pharmacometrics, such as analysing data from clinical trials ."

In 2003 he set up a working group called Monolix, bringing together researchers from Inria, Inserm (French National Institute for Health and Medical Research) and Inra (French National Institute for Agricultural Research) who specialised in statistics, biostatistics and pharmacology. This active group began to develop software, which was also called Monolix. "The pharmaceutical laboratories quickly showed interest in our work , Marc Lavielle affirms. It was notably the case of Johnson & Johnson, who enabled us to finance the recruitment of a post-doctoral student for two years. This gave serious impetus to the development of the Monolix software. "

Hand-in-hand with the pharmaceuticals industry

In 2007, Marc Lavielle was seconded to Inria Saclay, with the aim of reaching a new phase in the development of Monolix. "The Inria environment is much better suited to this type of project than a university environment ", Marc Lavielle points out. The interface with the industrial world is also one of Inria's key assets. In 2009, the DTI (Transfer and Innovation department) put into place a consortium of industrialists such as Novartis, Roche, Johnson & Johnson, Sanofi Aventis and, later, AstraZeneca. "These partners each provided €40,000 of funding per year to the project. As a result we had the necessary resources to really see our efforts come to fruition , Marc Lavielle sums up. Moreover, these partners gave us an industrial vision of our work, and contributed to our choice of areas for development. "

Two years later, in 2011, the Monolix software had been successfully completed, and was capable of practical applications in industry.  

When a research project becomes a company

"At this stage, Monolix was accessible via open source , remembers Marc Lavielle. However there was a growing number of support requests, and the industrialists using the software expressed the need for better visibility with regard to the future and the durability of the software. It was necessary to shift from the research world to the business world. "

This step was taken from 2011, with the creation of a spin-off called Lixoft, within the context of IncubAlliance, the technological incubator of the teaching and research establishments of the Paris-Saclay campus. Nevertheless, the originator of the Monolix project did not join the company himself, preferring to remain in his position at Inria. "Commercialising software is a different profession , Marc Lavielle insists. I wanted to continue doing what I loved - developing new methods and furthering our knowledge. Nonetheless, we rely on Lixoft to integrate these new ideas into the Monolix software and bring them to industry. From this perspective, Lixoft is our outlet, our essential link to the pharmaceutical world. " An approach that proved convincing, since Lixoft received the Oséo 2011 prize for the creation of innovative technology companies, in the "creation/development" category.

Research and industry: two inseparable cornerstones.

Besides the quality of his research work, what characterises Marc Lavielle is therefore his modern vision of research - a vision in which the academic world and the industrial world are two sides of the same coin. "Today in pharmacometrics, there is a real porosity between these two worlds , stresses Marc Lavielle. Many students do their theses in a university environment and then go on to work in industry. Inversely, some scientists come back to academic research from industry.  We need to make the best use of these opportunities for cross-fertilisation, so that it benefits innovation as much as possible. It is time for more statisticians to have an applied vision of their work and reach out to industry and towards putting their research into practice. "  


Donald E. Mager, Associate Professor of Pharmaceutical Sciences, University of Buffalo, President of the International Society of Pharmacometrics (ISoP)

"I first met Marc Lavielle when he was invited to share his insights and provide a series of mathematical modeling workshops at the University at Buffalo. Marc is a true innovator in the field of pharmacometrics. In his development of novel algorithms for conducting nonlinear mixed effects modeling, Marc’s outstanding research and implementation techniques have been transformative, providing robust and efficient methods for population-based modeling in diverse fields of study. The utilization of pharmacokinetic and pharmacodynamic systems analysis is a hallmark of model-informed drug development, and the advanced algorithms created by Marc and his colleagues directly address major prior limitations and enable population modeling with complex structural models. Marc is very passionate about mathematical modeling, a characteristic that comes through clearly in his lectures. He is a great professor, full of vision and expertise coupled with a genuine desire to facilitate student learning."