A hybrid mammalian cell cycle model
V. Noël, S. Vakulenko, O. Radulescu
Electronic Proceedings in Theoretical Computer Science 125 : 68-83, HSB 2013, Taormina

Hybrid modeling provides an effective solution to cope with multiple time scales dynamics in systems biology. Among the applications of this method, one of the most important is the cell cycle regulation. The machinery of the cell cycle, leading to cell division and proliferation, combines slow growth, spatio-temporal re-organisation of the cell, and rapid changes of regulatory proteins concentrations induced by post-translational modifications. The advancement through the cell cycle comprises a well defined sequence of stages, separated by checkpoint transitions. The combination of continuous and discrete changes justifies hybrid modelling approaches to cell cycle dynamics. We present a piecewise-smooth version of a mammalian cell cycle model, obtained by hybridization from a smooth biochemical model. The approximate hybridization scheme, leading to simplified reaction rates and binary event location functions, is based on learning from a training set of trajectories of the smooth model. We discuss several learning strategies for the parameters of the hybrid model.

 
Modèles réduits et hybrides de réseaux de réactions biochimiques: applications à la modélisation du cycle cellulaire
V. Noël
Thèse de doctorat. Université Rennes 1.

La modélisation des systèmes biologiques, particulièrement à l'échelle moléculaire, est une problématique nouvelle, issue de l'apport des techniques à haut débit. Le défi en modélisation mathématique est de pouvoir analyser le comportement de ces systèmes dynamiques de très grande dimension. L'enjeu est de taille, car la compréhension du fonctionnement normal et pathologique des cellules au niveau moléculaire, ouvre la voie aux thérapies ciblés pour des maladies systémiques telles que le cancer. Pour s'affranchir des problèmes liés à l'imprécision des valeurs des paramètres, cette thèse propose de travailler avec des ordres, plutôt qu'avec des valeurs précises de paramètres. Ceci conduit naturellement à l'utilisation de l'analyse tropicale pour obtenir des modèles réduits et hybrides. Ces développements ouvrent des nouvelles perspectives sur le plan mathématique, concernant l'étude de systèmes dynamiques. Cette étude propose quelques résultats concernant la tropicalisation des systèmes d'équations différentielles. Une autre partie de la thèse est consacrée à l'étude numérique des systèmes hybrides. La question ici est comment construire un modèle hybride qui reproduit un comportement expérimental donné, aussi comment identifier un modèle hybride à partir de séries temporelles. Cette thèse propose un algorithme original d'identification. Cet algorithme sépare le problème en deux sous-problèmes, notamment l'identification des paramètres des modes et l'identification des paramètres de commande des modes. Des applications à relativement grande échelle sont abordées par cette approche, notamment un modèle de cycle cellulaire chez les mammifères.

 
Tropicalization and Tropical equilibration for chemical kinetics
V. Noël, D. Grigoriev, S. Vakulenko, O. Radulescu
In Press, AMS Contemporary Mathematics, Tropical 2012, Moscow

Systems biology uses large networks of biochemical reactions to model the functioning of biological cells from the molecular to the cellular scale. The dynamics of dissipative reaction networks with many well separated time scales can be described as a sequence of successive equilibrations of different subsets of variables of the system. Polynomial systems with separation are equilibrated when at least two monomials, of opposite signs, have the same order of magnitude and dominate the others. These equilibrations and the corresponding truncated dynamics, obtained by eliminating the dominated terms, find a natural formulation in tropical analysis and can be used for model reduction.

 
Reduction of dynamical biochemical reaction networks in computational biology
O. Radulescu, A. N. Gorban, A. Zinovyev, V. Noël
Frontiers in Genetics 3, 131

Biochemical networks are used in computational biology, to model mechanistic details of systems involved in cell signaling, metabolism, and regulation of gene expression. Parametric and structural uncertainty, as well as combinatorial explosion are strong obstacles against analyzing the dynamics of large models of this type. Multiscaleness, an important property of these networks, can be used to get past some of these obstacles. s. Networks with many well separated time scales, can be reduced to simpler models, in a way that depends only on the orders of magnitude and not on the exact values of the kinetic parameters. The main idea used for such robust simplifications of networks is the concept of dominance among model elements, allowing hierarchical organization of these elements according to their effects on the network dynamics. This concept finds a natural formulation in tropical geometry. We revisit, in the light of these new ideas, the main approaches to model reduction of reaction networks, such as quasi-steady state and quasi-equilibrium approximations, and provide practical recipes for model reduction of linear and nonlinear networks. We also discuss the application of model reduction to the problem of parameter identification, via backward pruning machine learning techniques.

 
Hybrid models of the cell cycle molecular machinery
V. Noël, D. Grigoriev, S. Vakulenko, O. Radulescu
Electronic Proceedings in Theoretical Computer Science 92 : 88-105, HSB 2012, Newcastle

Piecewise smooth hybrid systems, involving continuous and discrete variables, are suitable models for describing the multiscale regulatory machinery of the biological cells. In hybrid models, the discrete variables can switch on and off some molecular interactions, simulating cell progression through a series of functioning modes. The advancement through the cell cycle is the archetype of such an organized sequence of events. We present an approach, inspired from tropical geometry ideas, allowing to reduce, hybridize and analyse cell cycle models consisting of polynomial or rational ordinary differential equations.

 
Tropical geometries and dynamics of biochemical networks. Application to hybrid cell cycle models
V. Noël, D. Grigoriev, S. Vakulenko, O. Radulescu
Electronic Notes in Theoretical Computer Science 284, pp. 75-91, SASB 2011, Venezia

We use the Litvinov-Maslov correspondence principle to reduce and hybridize networks of biochemical reactions. We apply this method to a cell cycle oscillator model. The reduced and hybridized model can be used as a hybrid model for the cell cycle. We also propose a practical recipe for detecting quasi-equilibrium QE reactions and quasi-steady state QSS species in biochemical models with rational rate functions and use this recipe for model reduction. Interestingly, the QE/QSS invariant manifold of the smooth model and the reduced dynamics along this manifold can be put into correspondence to the tropical variety of the hybridization and to sliding modes along this variety, respectively.

 
Algorithm for Identification of Piecewise Smooth Hybrid Systems : Application to Eukaryotic Cell Cycle Regulation
V. Noël, S. Vakulenko, O. Radulescu
Lecture Notes in Computer Science 6833, pp. 225-236, WABI 2011, Saakbrücken

We discuss piecewise smooth hybrid systems as models for regulatory networks in molecular biology. These systems involve both continuous and discrete variables. The discrete variables allow to switch on and off some of the molecular interactions in the model of the biological system. Piecewise smooth hybrid models are well adapted to approximate the dynamics of multiscale dissipative systems that occur in molecular biology. We show how to produce such models by a top down approach that use biological knowledge for a guided choice of important variables and interactions. Then we propose an algorithm for fitting parameters of the piecewise smooth models from data. We illustrate some of the possibilities of this approach by proposing hybrid versions of eukaryotic cell cycle regulation.

 
Piecewise smooth hybrid systems as models for networks in molecular biology
V. Noël, S. Vakulenko, O. Radulescu
Proceedings of JOBIM 2010, p57-62, Montpellier

We discuss piecewise smooth hybrid systems as models for regulatory networks in molecular biology. These systems involve both continuous and discrete variables. In the context of gene networks, the discrete variables allow to switch on and off some of the molecular interactions in the model of the biological system. Piecewise smooth hybrid models are well adapted to approximate the dynamics of multiscale dissipative systems that occur in molecular biology. We show how to produce such models by a top down approach that use biological knowledge for a guided choice of important variables and interactions. Then we propose an algorithm for fitting parameters of the piecewise smooth models from data. We illustrate some of the possibilities of this approach by proposing a minimal piecewise smooth model for the cell cycle