Constraint-based analysis of metabolism

Constraint-based analysis is a set of powerful techniques enabling the investigation of large systems of metabolic reactions for which no kinetic data are available. Relying on the topology and stoichiometry of metabolic networks, they enable the systematic enumeration of feasible reaction paths and define the boundaries of feasible flux distributions, yielding information about possible functional states of a cell.

Elementary flux modes are non-decomposable reaction paths that span the space of feasible steady-state flux distributions. We have been working on extending elementary mode analysis to investigate the dynamic activity of metabolic processes at the whole cell level. We developed algorithms for the decomposition of flux distributions onto individual elementary modes (Schwartz & Kanehisa, 2005), enabling quantitative biological interpretation of the activity of specific metabolic paths. We created methodologies for the integration of metabolic and transcription data (Nacher et al., 2006) and expanded elementary mode analysis to the genome scale, creating a framework to detect specific induced and repressed reaction paths and enabling transcription data to be associated to whole biological processes instead of individual genes (Schwartz et al., 2007).

We also work on methods to streamline the construction of genome-scale metabolic models and on theoretical approaches to understand the relationships between network structure and flux, particularly in cycles.

Dynamic modelling of large systems

Stoichiometric models provide limited insights into the functioning of cellular processes. To understand the detailed dynamics of cellular functions and their regulation, it is necessary to advance toward kinetic models where the behaviour of systems can be perturbed. The construction of genome-scale kinetic models is the next great challenge for systems biology. There is often inadequate knowledge of kinetic laws and associated parameter values, yet there is growing awareness that exact rate equations and parameters are often not crucial in determining the dynamic properties of large systems. We are developing the GRaPe software aimed at enabling the construction of large kinetic models based on generic rate equations.

Drug-metabolism interactions

Although investments by pharmaceutical companies have been growing continuously, the number of newly approved drugs has remained almost constant in the last decade. The traditional approach of drug development generally targets a single gene or gene product. However, many diseases are multifactorial which requires systemic effects of drug action to be taken into account. We have been working on characterising global interactions between drugs, metabolic networks and disease factors (Nacher & Schwartz, 2008; Nacher & Schwartz, 2012).

Systems modelling

We use a variety of modelling techniques ranging from logical networks to systems of ordinary differential equations in order to investigate specific biochemical pathways in collaboration with experimental investigators. Our current projects include:

• Glucocorticoid pathways to apoptosis (with Marija Krstic-Demonacos).
• Temperature compensation of the Neurospora circadian clock (with Sue Crosthwaite).
• Thermodynamic modelling of yeast metabolism (with Daniela Delneri).
• The cAMP pathway in the pathogenic yeast Candida albicans (with Lubomira Stateva).
• Effects of environmental factors on plant metabolism (with Giles Johnson).