Showing papers presented at "Computational Methods in Systems Biology in 2004"

Formal molecular biology

Abstract: A language of formal proteins, the K-calculus, is introduced. Interactions are modeled at the domain level, bonds are represented by means of shared names, and reactions are required to satisfy a causality requirement of monotonicity.An example of a simplified signalling pathway is introduced to illustrate how standard biological events can be expressed in our protein language. A more comprehensive example, the lactose operon, is also developed, bringing some confidence in the formalism considered as a modeling language.Then a finer-grained concurrent model, the mK-calculus, is considered, where interactions have to be at most binary. We show how to embed the coarser-grained language in the latter, a properly which we call self-assembly.Finally we show how the finer-grained language can itself be encoded in π-calculus, a standard foundational language for concurrency theory.

BioAmbients: an abstraction for biological compartments

Abstract: Biomolecular systems, composed of networks of proteins, underlie the major functions of living cells. Compartments are key to the organization of such systems. We have previously developed an abstraction for biomolecular systems using the π-calculus process algebra, which successfully handled their molecular and biochemical aspects, but provided only a limited solution for representing compartments. In this work, we extend this abstraction to handle compartments. We are motivated by the ambient calculus, a process algebra for the specification of process location and movement through computational domains. We present the BioAmbients calculus, which is suitable for representing various aspects of molecular localization and compartmentalization, including the movement of molecules between compartments, the dynamic rearrangement of cellular compartments, and the interaction between molecules in a compartmentalized setting. Guided by the calculus, we adapt the BioSpi simulation system, to provide an extended modular framework for molecular and cellular compartmentalization, and we use it to model and study a complex multi-cellular system.

Brane calculi

Abstract: We introduce a family of process calculi with dynamic nested membranes. In contrast to related calculi, including some developed for biological applications, active entities here are tightly coupled to membranes, and can perform interactions on both sides of a membrane. That is, computation happens on the membrane, not inside of it.

Beta binders for biological interactions

Abstract: This paper presents binders and operators, in the process calculi tradition, to reason about biological interactions. Special binders are added to wrap a process just as membranes enclose some living matter and hence to mimick biological interfaces. A few operators are then added to the pi-calculus kernel to describe the dynamics of those interfaces.

Modeling and querying biomolecular interaction networks

Abstract: We introduce a formalism to represent and analyze protein-protein and protein-DNA interaction networks. We illustrate the expressivity of this language, by proposing a formal counterpart of Kohn's compilation on the mammalian cell-cycle control. This effectively turns an otherwise static knowledge into a discrete transition system incorporating a qualitative description of the dynamics. We then propose to use the computation tree logic (CTL) as a query language for querying the possible behaviors of the system. We provide examples of biologically relevant queries expressed in CTL about the mammalian cell-cycle control and show the effectiveness of symbolic model checking tools to evaluate CTL queries in this context.

The biochemical abstract machine BIOCHAM

Abstract: In this article we present the Biochemical Abstract Machine BIOCHAM and advocate its use as a formal modeling environment for networks biology. Biocham provides a precise semantics to biomolecular interaction maps. Based on this formal semantics, the Biocham system offers automated reasoning tools for querying the temporal properties of the system under all its possible behaviors. We present the main features of Biocham, provide details on a simple example of the MAPK signaling cascade and prove some results on the equivalence of models w.r.t. their temporal properties.

Modelling biochemical pathways through enhanced π-calculus

Abstract: We use the π-calculus to model the evolution of biochemical systems, taking advantage of their similarities with global computation applications. First, we present a reduction semantics for the π-calculus from which causality and concurrency can be mechanically derived. We prove that our semantics agrees with the causal definitions presented in the literature. We also extend our semantics to model biological compartments. Then, we show the applicability of our proposal on a couple of biological examples.

Projective brane calculus

Abstract: A refinement of Cardelli’s brane calculus [1] is introduced where membrane actions are directed. This modification brings the language closer to biological membranes and also obtains a symmetric set of membrane interactions. An associated structural congruence, termed the projective equivalence, is defined and shown to be preserved under all possible system evolutions. Comparable notions of projective equivalence can be developed in other hierarchical process calculi and might be of interest in other applications.

Taming the complexity of biochemical models through bisimulation and collapsing: theory and practice

Abstract: Many biological systems can be modeled using systems of ordinary differential algebraic equations (e.g., S-systems), thus allowing the study of their solutions and behavior automatically with suitable software tools (e.g., PLAS, Octave/Matlabtm). Usually, numerical solutions (traces or trajectories) for appropriate initial conditions are analyzed in order to infer significant properties of the biological systems under study. When several variables are involved and the traces span over a long interval of time, the analysis phase necessitates automation in a scalable and efficient manner. Earlier, we have advocated and experimented with the use of automata and temporal logics for this purpose (XS-systems and Simpathica) and here we continue our investigation more deeply.We propose the use of hybrid automata and we discuss the use of the notions of bisimulation and collapsing for a "qualitative" analysis of the temporal evolution of biological systems. As compared with our previous approach, hybrid automata allow maintenance of more information about the differential equations (S-system) than standard automata. The use of the notion of bisimulation in the definition of the projection operation (restrictions to a subset of "interesting" variables) makes it possible to work with reduced automata satisfying the same formulae as the initial ones. Finally, the notion of collapsing is introduced to move toward still simpler and equivalent automaton taming the complexity in terms of states whose number depends on the attained level of approximation.

VICE: a VIrtual CEll

Abstract: We report on the specification and analysis of VICE, a hypothetical cell with a genome as basic as possible We used an enhanced version of the π-calculus and a prototype running it to study the behaviour of VICE The results of our experimentation in silico confirm that our virtual cell “survives” in an optimal environment and shows a behaviour similar to that of real prokaryotes

Combining state-based and scenario-based approaches in modeling biological systems

Abstract: Biological systems have recently been shown to share many of the properties of reactive systems. This observation has led to the idea of using methods devised for the construction (engineering) of complex reactive systems to the modeling (reverse-engineering) of biological systems, in order to enhance biological comprehension. Here we suggest to combine the two formal approaches used in our group — the state-based formalism of statecharts and the scenario-based formalism of live sequence charts (LSCs). We propose that biological observations are better formalized in the form of LSCs, while biological mechanistic models would be more natural to specify using statecharts. Combining the two approaches would enable one to verify the proposed mechanistic models against the real data. The biological observations can be compared to the requirements in an engineered system, and the mechanistic model would be analogous to the implementation. While requirements are used to design an implementation, here the observations are used to motivate the invention of the mechanistic model. In both cases consistency of one with the other must be established, by testing or by formal verification.

Modeling the molecular network controlling adhesion between human endothelial cells: inference and simulation using constraint logic programming

Abstract: Cell-cell adhesion plays a critical role in the formation of tissues and organs. Adhesion between endothelial cells is also involved in the control of leukocyte migration across the endothelium of blood vessels. The most important players in this process are probably identified and the overall organization of the biochemical network can be drawn, but knowledge about connectivity is still incomplete, and the numerical values of kinetic parameters are unknown. This calls for qualitative modeling methods. Our aim in this paper is twofold: (i) to integrate in a unified model the biochemical network and the genetic circuitry. For this purpose we transform our system into a system of piecewise linear differential equations and then use Thomas theory of discrete networks. (ii) to show how constraints can be used to infer ranges of parameter values from observations and, with the same model, perform qualitative simulations.

Model checking biological systems described using ambient calculus

Abstract: We propose a way of performing model checking analysis for biological systems. The technics were developed for a CTL* logic built upon Ambient Calculus. We introduce labeled syntax trees for ambient processes and use them as possible worlds in a Kripke structure developed for a propositional branching temporal logic. The accessibility relation over labeled syntax trees is generated by the reduction over corresponding Ambient Calculus processes. Providing the algorithms for calculating the accessibility relation between states, we open the perspective of using model checking algorithms developed for temporal logics in analyzing any phenomena described in Ambient Calculus.

General stochastic hybrid method for the simulation of chemical reaction processes in cells

Abstract: Stochastic approaches are required for the simulation of biochemical systems like signal transduction networks, since high fluctuations and extremely low particle numbers of some species are ubiquitous. Computational problems arise from the huge differences among the timescales on which the reactions occur, causing high cost for stochastically exact simulations. Here, we demonstrate a general hybrid method combining the exact Gillespie algorithm with a system of stochastic differential equations (SDEs). This technique provides a smooth and correct transition between subsets of ’slow’ and ’fast’ reactions instead of abruptly cutting the stochastic effects above a certain particle number. The method was successfully applied to mitochondrial Cytochrome C release in the CD95-induced apoptosis pathway. Moreover, this approach can also be used for other kinds of Markov processes.

Modelling metabolic pathways using stochastic logic programs-based ensemble methods

Abstract: In this paper we present a methodology to estimate rates of enzymatic reactions in metabolic pathways. Our methodology is based on applying stochastic logic learning in ensemble learning. Stochastic logic programs provide an efficient representation for metabolic pathways and ensemble methods give state-of-the-art performance and are useful for drawing biological inferences. We construct ensembles by manipulating the data and driving randomness into a learning algorithm. We applied failure adjusted maximization as a base learning algorithm. The proposed ensemble methods are applied to estimate the rate of reactions in metabolic pathways of Saccharomyces cerevisiae. The results show that our methodology is very useful and it is effective to apply SLPs-based ensembles for complex tasks such as modelling of metabolic pathways.

A multi-scale constraint programming model of alternative splicing regulation

Abstract: Alternative splicing is a key process in post-transcriptional regulation, by which different mature RNA can be obtained from the same premessenger RNA. The resulting combinatorial complexity contributes to biological diversity, especially in the case of the human immunodeficiency virus HIV-1. Using a constraint programming approach, we develop a model of the alternative splicing regulation in HIV-1. Our model integrates different scales (single site vs. multiple sites), and thus allows us to exploit several types of experimental data available to us.

Towards reusing model components in systems biology

Abstract: For reusing model components, it is crucial to understand what information is needed and how it should be presented. The centrality of abstraction being inherent in the modelling process distinguishes model components from software components and makes their reuse even more difficult. Objectives and assumptions which are often difficult to explicitate become an important aspect in describing model components. Following the argumentation line of the Web Service ontology OWL-S, we propose a set of metadata which is structured into profile, process model, and grounding to describe model components. On the basis of the specific model component Tryptophan Synthase, its metadata is refined in XML. The reuse of the described model component is illustrated by integrating it into a model of the Tryptophan operon.

Residual bootstrapping and median filtering for robust estimation of gene networks from microarray data

Abstract: We propose a robust estimation method of gene networks based on microarray gene expression data. It is well-known that microarray data contain a large amount of noise and some outliers that interrupt the estimation of accurate gene networks. In addition, some relationships between genes are nonlinear, and linear models thus are not enough for capturing such a complex structure. In this paper, we utilize the moving boxcel median filter and the residual bootstrap for constructing a Bayesian network in order to attain robust estimation of gene networks. We conduct Monte Carlo simulations to examine the properties of the proposed method. We also analyze Saccharomyces cerevisiae cell cycle data as a real data example.

Biomimetic in silico devices

Abstract: We introduce biomimetic in silico devices, and means for validation along with methods for testing and refining them. The devices are constructed from adaptable software components designed to map logically to biological components at multiple levels of resolution. In this report we focus on the liver; the goal is to validate components that mimic features of the lobule (the hepatic primary functional unit) and dynamic aspects of liver behavior, structure, and function. An assembly of lobule-mimetic devices represents an in silico liver. We validate against outflow profiles for sucrose administered as a bolus to isolated, perfused rat livers. Acceptable in silico profiles are experimentally indistinguishable from those of the in situ referent. This new technology is intended to provide powerful new tools for challenging our understanding of how biological functional units function in vivo.

CMBSlib: a library for comparing formalisms and models of biological systems

Abstract: We present CMBSlib, a library of Computational Models of Biological Systems. It is aimed at providing a list of test problems for formalisms, modeling issues and implementation issues in systems biology. The main motivation for CMBSlib is to stimulate research on the formal modeling of biological systems, by facilitating the exchange of formal models between researchers, and by providing a forum of comparison and validation of not only models, but also modeling formalisms and implementations. Unlike a standardization effort, CMBSlib welcomes the most exotic formalisms and models provided they attack the modeling of well documented biological systems. Models of biological systems written in any referenced formalism can be submitted to CMBSlib. No special format or standard is required. We discuss the advantages of and problems encountered in building such a library, give an example of typical entry in the library, and most of all we invite the community to become active contributors to CMBSlib.

Developing SBML beyond level 2: proposals for development

Abstract: The Systems Biology Markup Language (SBML) is an XML-based exchange format for computational models of biochemical networks. SBML Level 2, whose definition was established in June 2003, includes several enhancements to the original Level 1. This paper includes a brief overview of Level 2. Several proposals are under development to extend SBML to create Level 3. These include diagrams, 2-D and 3-D spatial characteristics, arrays, model composition and multi-component chemical species. This paper describes the current proposals for the last two features.

IMGT-Choreography: processing of complex immunogenetics knowledge

Abstract: IMGT, the international ImMunoGeneTics information system® (http://imgt.cines.fr) is a high quality integrated knowledge resource specialized in immunoglobulins (IG), T cell receptors (TR), major histocompatibility complex (MHC) and related proteins of the immune system (RPI) of human and other vertebrates. IMGT provides a common access to standardized data from genome, proteome, genetics and tridimensional structures. The accuracy and the consistency of IMGT data are based on IMGT-ONTOLOGY, a semantic specification of terms used in immunogenetics and immunoinformatics. IMGTONTOLOGY is formalized using XML schemas (IMGT-ML) for interoperability with other information systems. We are developing Web services to automatically query IMGT databases and tools. This is the first step towards IMGT-Choreography which will trigger and coordinate dynamic interactions between IMGT Web services, to process complex significant biological and clinical requests. IMGT-Choreography will further increase the IMGT leadership in immunogenetics medical research (repertoire analysis in autoimmune diseases, AIDS, leukemias,...), biotechnology related to antibody engineering and therapeutical approaches.

Graph-Based modeling of biological regulatory networks: introduction of singular states

Abstract: In the field of biological regulation, models extracted from experimental works are usually complex networks comprising intertwined feedback circuits. R. Thomas and coworkers introduced a qualitative description of the dynamics of such regulatory networks, called the generalized logical analysis, and used the concept of circuit-characteristic states to identify all steady states and functional circuits. These characteristic states play an essential role on the dynamics of the system, but they are not represented in the state graph. In this paper we present an extension of this formalism in which all singular states including characteristic ones are represented. Consequently, the state graph contains all steady states. Model checking is then able to verify temporal properties concerning singular states. Finally, we prove that this new modeling is coherent with R. Thomas’ modeling since all paths of R. Thomas’ dynamics are represented in the new state graph, which in addition shows the influence of singular states on the dynamics.

Biological domain identification based in codon usage by means of rule and tree induction

Abstract: There are three domains in living nature: archaea, bacteria and eukarya. It has been shown, trough a number of multivariate tools, that codon usage, a 64 dimensional vector that stablishes how often a given organism makes use of each codon, is related to domain. Another method is proposed here based in rule and tree induction from codon usage of several organisms. It is shown that domain can be identified trough codon usage and a simple set of rules. Two methods were applied, CN 2 and C 4.5. Obtained rules describe data better than other methods, in the sense that are topological interpretable and have phenomenological meaning.

Autonomous mobile robot control based on white blood cell chemotaxis

Abstract: This paper presents a biologically inspired algorithm to control an autonomous robot tracking a target. The algorithm is designed to mimic the behavior of a human neutrophil, a type of white blood cell that travels to sites of infection and digests bacterial antagonists. Neutrophils are known to be highly sensitive to low levels of chemical stimuli, robust to noise, and are capable navigating unknown terrain, all qualities that would be desired in an autonomous robot. In this paper we model a neutrophil as a collaborative control system, demonstrate the robustness of this algorithm, and suggest a computationally cheap method of implementation. Our simulations show that the performance of the robot is unaffected by constant disturbances and it is robust to random noise levels up to 5 times the tracking signal. Additionally, we demonstrate that this algorithm, as well the current models of neutrophil chemotaxis, are equivalent to a sensor fusion problem that optimizes directional sensing in the presence of noise.

An explicit upper bound for the approximation ratio of the maximum gene regulatory network problem

Abstract: One of the combinatorial models for the biological problem of inferring gene regulation networks is the Maximum Gene Regulatory Network Problem, shortly MGRN, proposed in [2]. The problem is NP-hard [2], consequently the attention has shifted towards approximation algorithms, leading to a polynomial-time 1/2-approximation algorithm [2], while no upper bound on the possible approximation ratio was previously known. In this paper we make a first step towards closing the gap between the best known and the best possible approximation factors, by showing that no polynomial-time approximation algorithm can have a factor better than 1 – (1/8) / (1+e2) unless RP=NP.

Building and analysing an integrative model of HIV-1 RNA alternative splicing

Abstract: We present a multi-site model describing the alternative use of the RNA splicing sites A3, A4, A5 and A7 in the human immunodeficiency virus HIV-1. Our goal is to integrate experimental data obtained on individual splicing sites into a global model of HIV-1 RNA alternative splicing. We give a qualitative validation of our model, and analyse the possible impact of variations of regulatory protein concentrations on virus multiplication.

Spatial modeling and simulation of diffusion in nuclei of living cells

Abstract: The mobility of fluorescently labelled molecules in the interphase nucleus has been increasingly employed to investigate the spatial organization of the interchromosomal space. We suggest an improved two-dimensional anisotropic diffusion model to address the inhomogeneous nature of nuclear organization, which is at odds with the generally applied ’well-mixed’ compartmental assumption. To consider the transfer function of the imaging system, we derived a modified fundamental solution of the two-dimensional, time-dependent diffusion equation. The model was validated through comparison of the forward simulation results with fluorescence recovery after photobleaching experiments using nuclear localization signal (NLS) – tagged YFP recorded by confocal laser scanning microscopy. To improve the fit error in the vicinity of the nuclear boundary, we suggest an isotropic diffusion model with Neumann boundary condition accounting for the exact shape of the nuclear boundary. The suggested approach is a first step towards diffusion tomography of the cell nucleus.

Black box checking for biochemical networks

Abstract: We propose black box checking as a framework for analyzing biochemical networks. Black box checking was originally introduced by Peled, Yannakakis and Vardi in the context of formal verification of concurrent systems as a strategy that combines model checking and testing, as two main techniques in that area. Based on the natural analogy between biochemical networks and concurrent systems we argue that black box checking can be used to design and perform experiments in a systematic manner, and also to learn about the network underlying mechanisms. We also discuss potential applications with emphasis on forward engineering of biochemical networks.

The biodegradation network, a new scenario for computational systems biology research

Abstract: The study of the global properties of complex metabolic networks in organisms such as E. coli and Yeast has become a priority in the new area of Computational Systems Biology, leading to the first models about the organization and evolution of these networks. Here we propose the analysis of biodegradation networks as an alternative to this, now classical, studies of metabolic networks. The biodegradation reactions carried out by communities of bacteria in contaminated environments offer interesting computational and biological advantages and challenges. Here we first describe our preliminary results on the analysis of the systems properties of the biodegradation network and the comparison with the other metabolic networks, and second describe the computational developments necessary for the analysis of the information. In particular we introduce the MetaRouter system, as an integrated information system for Biodegradation. Furthermore we describe the current work for improving the system with an alternative enzyme-centric view and new rule-based systems for the prediction of the capacity of biological systems to degrade new chemical compounds.