Blanc SIMI 2 - Sciences de l'information, de la matière et de l'ingénierie : Sciences de l’information, simulation

Langage, time representations and hybrid models for the analysis of incomplete models in molecular biology – BioTempo

Submission summary

Systems biology has been growing very quickly for the last ten years and it is now reaching a cornerstone: is it formed by two sub-domains that are quite independent. The first sub-domain of systems biology is sometimes referred to as integrative biology. It aims at formalizing and annotating data that are related to cellular interactions or observations, and to extract a signal by using probabilistic approaches (Bayesian analyses, support vector machines...). Such analyses have led to network reconstruction and even to experimental design. However, they require very large scale of data that are often accessible only for well known species. The second sub-domain may be called dynamical modelling. It refers to the design of computational or mathematical models of interactions within the cell, in order to simulate the behaviour of a biological system and predict its main features. This domain belongs to both mathematics and computer science (depending on the discreteness or variable representation). It requires a fine tuning of the parameters obtained from spatio-temporal observations that are not (yet) available by large-scale observation data studied in integrative biology. Properties are obtained by sensitivity analyses, robustness or model checking.

With regards to expectations of biologists, the distance between integrative biology and dynamical modelling is now the main weakness of systems biology. To fill this gap, we wish to focus on new computational and mathematical concepts for models. From a theoretical point of view, the main difference between integrative biology and dynamical modelling lies in the integration of time in the study. Integrative biology looks for significant signals in data of different origins; time-series exist but they are mainly used for correlation purposes. Dynamical modelling, however, aims at describing trajectories – that is, the evolution of products with respect to time – with the best available precision; therefore additional information on protein bindings, global variation levels or knock-out effects are not considered.

The panel of modelling approaches that are currently available share a main characteristic : the representation of time is correlated with the representation of variables (product activities). Roughly, models can be static (time is not considered, this is the core of integrative biology), chronological (time and variables are discrete, leading to model checking) or chronometric (time and variable are continuous, leading to differential equations). For computational and parameter estimation purposes, the size of models strongly decreases when switching from static to chronological models. Our goal is to combine and complete these approaches (static, chronological and chronometric) in order to gain in the analysis and interpretation of biological data.

In particular, we wish to disjoint the representations of time and variables. For that purpose, we have gathered computer scientists specialized in graph algorithms, programming logics, model checking (including hybrid models), and Markov chain, together with mathematicians specialized in the modelling of incompletely described biological systems. All participants have worked on concrete biological applications so that they know the limitations of currently available methods. Biologists teams will also join the project, allowing us to validate the new formalisms that will be developed over three biological system (copper bacteria, sea urchin translation, TFGbeta signalling). We will use already produced data and preliminary models to test our approaches. This would allow to to focus on computational and mathematical challenges raised by these new hybrid models for biological systems, meanwhile giving us the opportunity to compare our respective approaches on common biological applications.

Project coordination

Anne Siegel (CNRS - DELEGATION REGIONALE BRETAGNE ET PAYS- DE-LA-LOIRE) – anne.siegel@irisa.fr

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

CNRS / I3S CNRS - DELEGATION REGIONALE COTE D'AZUR
CNRS / IRISA CNRS - DELEGATION REGIONALE BRETAGNE ET PAYS- DE-LA-LOIRE
LINA UNIVERSITE DE NANTES

Help of the ANR 310,000 euros
Beginning and duration of the scientific project: - 36 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter