Research

Organization principles of biochemical networks

The main goal of this research line is to discover, understand and exploit general rules that (a) relate the design (i.e. naturally evolved molecular mechanisms) of biochemical systems to their function, and (b) hold across processes, cell types and organisms. We envisage that these network-structure / function relationships will play in biomedicine and bioengineering a role analogous to that of QSAR in pharmacology. Current objects of interest are redox signaling, antioxidant defense and metabolic networks. Using mathematical modeling, systems analysis and bioinformatics tools, we have identified recurrent structural and functional motifs in all these biomolecular networks and derived design principles (relationships among kinetic parameters and component concentrations) that these motifs must fulfill so that they perform their function adequately. These predictions are thoroughly supported by experimental observations in a variety of organisms and permitted rationalizing the phenotypes of mutations and stress responses. We are working towards exploring translational implications of these design/function relationships in degenerative diseases and vascular development. In parallel, we are developing novel experimental (fluxomics and synthetic biology) methodologies to determine critical parameters in these applications. 

Modeling permeation through physiological barriers

The long-term goal of this research line is to develop quantitative structure-activity relationships (QSAR) for the permeation of physiological barriers by drugs. Namely tight endothelia such as the blood-brain barrier (BBB). Failure to cross the BBB is the main factor of attrition in the development of psycho-active drugs, and is causing some of the main pharmaceutical corporations to abandon the development of such drugs altogether. The bioavailability of xenobiotics at the brain is strongly affected by their interaction with lipid bilayers and blood components (albumin, lipoproteins, erythrocytes and membranes of endothelial cells). Our work shows that the partition of drugs among the compartments strongly affects the timing and effectiveness of their permeation across the BBB. Using computational tools ranging from molecular dynamics to kinetic modeling and experimental data supplied by our collaborators, we are working towards taking various important aspects of the topology of the BBB into account so that more-accurate predictions can be made.

Computational tools for biomolecular systems.

The main goal of this research line is to develop effective computational tools to analyze complex biomolecular systems and reaction networks. Namely, in support of the activities of the research lines described above. Developments range from very fundamental computer-science methods to speed-up numerical computation in a broad range of applications, to tools for reaction networks. The latter have two focus areas. (a) Tools for characterizing the relationship between design and performance of biomolecular reaction networks. (b) A rule-based modeling approach to construct and analyze very complex reaction networks, taking the stereochemistry of the compounds into account. This approach is being applied to analyze the autoxidation of polyunsaturated fatty acids leading to the formation of reactive aldehydes (e.g. hydroxynonenal) that are very relevant in chronic and degenerative diseases and in food biotechnology.