Broadly speaking, I am interested in the application of concepts and methods of engineering to study the functioning of living cells. In particular, I aim to elucidate the processing of information in human cells, and uncovering the molecular basis of the deregulation of these processes in a variety of diseases.

Computational analysis of signaling networks in normal and diseased cells 

Human cells are equipped with exquisite sensing systems that allow them to receive and process the information encoded in myriad extracellular stimuli. These sensing systems are embedded in very complex signaling networks. Understanding how these networks work is a fascinating academic challenge, but also of great practical importance, since alterations in the functioning of these networks underlies the development of diseases such as cancer or diabetes. Considerable effort has been devoted to identify proteins that can be targeted to reverse this deregulation, but their benefit is often not the expected one: it is hard to assess their influence on the signaling network as a whole and thus their net effect on the behavior of the diseased cell. Such a global understanding can only be achieved by a combination of experimental and computational analysis. 

During my Ph.D. at the Max-Planck-Institute for the Dynamics of Complex Technical Systems, I worked, under the supervision of Prof. E. D. Gilles, on a number of theoretical questions about the structure and functioning of signaling networks, based on the concept of modularity, as explain in the section Modular analysis of signal transduction networks.

In my postdoctoral training in the Department of Systems Biology at Harvard Medical School (group of Prof. Peter Sorger ) and the Department of Biological Engineering at MIT (group of Prof. Douglas Lauffenburger),  we develop methods mathematical models that integrate high-throughput biochemical data with various sources of prior knowledge, with an emphasis on providing both predictive power of new experiments and insight on the functioning of the signaling network (Saez-Rodriguez, Alexopoulos, et al., Mol. Syst. Biol., 2009).  We then use these models to better understand how signaling is altered in human disease and predict effective therapeutic targets.  As a case study, we have focused on analyzing the differences between primary and transformed hepatocytes (liver cells), being able to uncover significant differences in the rewiring of their signaling networks.

Productive integration of data and computation requires an effective workflow that pulls together all the steps that link experiments to mathematical models and analysis. We are developing a platform to facilitate this process by creating a set of interoperable software tools incorporating public standards. Currently two modules are available: DataRail, a toolbox for managing, transforming and visualizing experimental data (Saez-Rodriguez, Goldsipe, et al., Bioinformatics, 2008), and CellNetOptimizer, to create logic-based models based on experimental data sets and prior knowledge (Saez-Rodriguez, Alexopoulos, et al., Mol. Syst. Biol., 2009).

In our group, recently started at the  European Bioinformatics Institute (EBI-EMBL, part of the European Molecular Biology Laboratory), we will continue working on the development and application of computational methods to understand the disregulation  of cellular information processing in disease. You can find more information here.

I am also involved in a community effort to advance the inference of mathematical models of cellular networks: DREAM (Dialogue for Reverse Engineering Assessments and Methods). DREAM is articulated around "challenges" posed to the community, and discussion of the results in an annual meeting. The results and methods are then shared with the community.