The systems biology markup language (SBML): a medium for representation and exchange of biochemical network models

Toward large-scale modeling of the microbial cell for computer simulation

In the post-genomic era, the large-scale, systematic, and functional analysis of all cellular components using transcriptomics, proteomics, and metabolomics, together with bioinformatics for the analysis of the massive amount of data generated by these "omics" methods are the focus of intensive research activities. As a consequence of these developments, systems biology, whose goal is to comprehend the organism as a complex system arising from interactions between its multiple elements, becomes a more tangible objective. Mathematical modeling of microorganisms and subsequent computer simulations are effective tools for systems biology, which will lead to a better understanding of the microbial cell and will have immense ramifications for biological, medical, environmental sciences, and the pharmaceutical industry. In this review, we describe various types of mathematical models (structured, unstructured, static, dynamic, etc.), of microorganisms that have been in use for a while, and others that are emerging. Several biochemical/cellular simulation platforms to manipulate such models are summarized and the E-Cell system developed in our laboratory is introduced. Finally, our strategy for building a "whole cell metabolism model", including the experimental approach, is presented.

Quantitative information management for the biochemical computation of cellular networks

Understanding complex protein networks within cells requires the ability to develop quantitative models and to numerically compute the properties and behavior of the networks. To carry out such computational analysis, it is necessary to use modeling tools and information management systems (IMSs) where the quantitative data, associated to its biological context, can be stored, curated, and reliably retrieved. We have focused on the biochemical computation of cellular interactions and developed an IMS that stores both quantitative information on the cellular components and their interactions, and the basic reactions governing those interactions. This information can be used to construct pathways and eventually large-scale networks. This system, SigPath, is available on the Internet (http://www.sigpath.org). Key features of the approach include (i) the use of background information (for example, names of molecules, aliases, and accession codes) to ease data submission and link this quantitative database with other qualitative databases, (ii) a strategy to allow refinement of information over time by multiple users, (iii) the development of a data representation that stores both qualitative and quantitative information, and (iv) features to assist contributors and users in assembling custom quantitative models from the information stored in the IMS. Currently, models assembled in SigPath can be automatically exported to several computing environments, such as Kinetikit/Genesis, Virtual Cell, Jarnac/JDesigner, and JSim. We anticipate that, when appropriately populated, such a system will be useful for large-scale quantitative studies of cell-signaling networks and other cellular networks. SigPath is distributed under the GNU General Public License.

PIML: the Pathogen Information Markup Language

MOTIVATION

A vast amount of information about human, animal and plant pathogens has been acquired, stored and displayed in varied formats through different resources, both electronically and otherwise. However, there is no community standard format for organizing this information or agreement on machine-readable format(s) for data exchange, thereby hampering interoperation efforts across information systems harboring such infectious disease data.

RESULTS

The Pathogen Information Markup Language (PIML) is a free, open, XML-based format for representing pathogen information. XSLT-based visual presentations of valid PIML documents were developed and can be accessed through the PathInfo website or as part of the interoperable web services federation known as ToolBus/PathPort. Currently, detailed PIML documents are available for 21 pathogens deemed of high priority with regard to public health and national biological defense. A dynamic query system allows simple queries as well as comparisons among these pathogens. Continuing efforts are being taken to include other groups' supporting PIML and to develop more PIML documents.

AVAILABILITY

All the PIML-related information is accessible from http://www.vbi.vt.edu/pathport/pathinfo/

CellML: its future, present and past

Advances in biotechnology and experimental techniques have lead to the elucidation of vast amounts of biological data. Mathematical models provide a method of analysing this data; however, there are two issues that need to be addressed: (1) the need for standards for defining cell models so they can, for example, be exchanged across the World Wide Web, and also read into simulation software in a consistent format and (2) eliminating the errors which arise with the current method of model publication. CellML has evolved to meet these needs of the modelling community. CellML is a free, open-source, eXtensible markup language based standard for defining mathematical models of cellular function. In this paper we summarise the structure of CellML, its current applications (including biological pathway and electrophysiological models), and its future development--in particular, the development of toolsets and the integration of ontologies.

Pathway databases

Network representations of biological pathways offer a functional view of molecular biology that is different from and complementary to sequence, expression, and structure databases. There is currently available a wide range of digital collections of pathway data, differing in organisms included, functional area covered (e.g., metabolism vs. signaling), detail of modeling, and support for dynamic pathway construction. While it is currently impossible for these databases to communicate with each other, there are several efforts at standardizing a data exchange language for pathway data. Databases that represent pathway data at the level of individual interactions make it possible to combine data from different predefined pathways and to query by network connectivity. Computable representations of pathways provide a basis for various analyses, including detection of broad network patterns, comparison with mRNA or protein abundance, and simulation.

From the sequence to cell modeling: comprehensive functional genomics in Escherichia coli

As a result of the enormous amount of information that has been collected with E. coli over the past half century (e.g. genome sequence, mutant phenotypes, metabolic and regulatory networks, etc.), we now have detailed knowledge about gene regulation, protein activity, several hundred enzyme reactions, metabolic pathways, macromolecular machines, and regulatory interactions for this model organism. However, understanding how all these processes interact to form a living cell will require further characterization, quantification, data integration, and mathematical modeling, systems biology. No organism can rival E. coli with respect to the amount of available basic information and experimental tractability for the technologies needed for this undertaking. A focused, systematic effort to understand the E. coli cell will accelerate the development of new post-genomic technologies, including both experimental and computational tools. It will also lead to new technologies that will be applicable to other organisms, from microbes to plants, animals, and humans. E. coli is not only the best studied free-living model organism, but is also an extensively used microbe for industrial applications, especially for the production of small molecules of interest. It is an excellent representative of Gram-negative commensal bacteria. E. coli may represent a perfect model organism for systems biology that is aimed at elucidating both its free-living and commensal life-styles, which should open the door to whole-cell modeling and simulation.

Modeling regulatory networks at Virginia Tech

The life of a cell is governed by the physicochemical properties of a complex network of interacting macromolecules (primarily genes and proteins). Hence, a full scientific understanding of and rational engineering approach to cell physiology require accurate mathematical models of the spatial and temporal dynamics of these macromolecular assemblies, especially the networks involved in integrating signals and regulating cellular responses. The Virginia Tech Consortium is involved in three specific goals of DARPA's computational biology program (Bio-COMP): to create effective software tools for modeling gene-protein-metabolite networks, to employ these tools in creating a new generation of realistic models, and to test and refine these models by well-conceived experimental studies. The special emphasis of this group is to understand the mechanisms of cell cycle control in eukaryotes (yeast cells and frog eggs). The software tools developed at Virginia Tech are designed to meet general requirements of modeling regulatory networks and are collected in a problem-solving environment called JigCell.

From annotated genomes to metabolic flux models and kinetic parameter fitting

Significant advances in system-level modeling of cellular behavior can be achieved based on constraints derived from genomic information and on optimality hypotheses. For steady-state models of metabolic networks, mass conservation and reaction stoichiometry impose linear constraints on metabolic fluxes. Different objectives, such as maximization of growth rate or minimization of flux distance from a reference state, can be tested in different organisms and conditions. In particular, we have suggested that the metabolic properties of mutant bacterial strains are best described by an algorithm that performs a minimization of metabolic adjustment (MOMA) upon gene deletion. The increasing availability of many annotated genomes paves the way for a systematic application of these flux balance methods to a large variety of organisms. However, such a high throughput goal crucially depends on our capacity to build metabolic flux models in a fully automated fashion. Here we describe a pipeline for generating models from annotated genomes and discuss the current obstacles to full automation. In addition, we propose a framework for the integration of flux modeling results and high throughput proteomic data, which can potentially help in the inference of whole-cell kinetic parameters.

MathSBML: a package for manipulating SBML-based biological models

MathSBML is a Mathematica package designed for manipulating Systems Biology Markup Language (SBML) models. It converts SBML models into Mathematica data structures and provides a platform for manipulating and evaluating these models. Once a model is read by MathSBML, it is fully compatible with standard Mathematica functions such as NDSolve (a differential-algebraic equations solver). MathSBML also provides an application programming interface for viewing, manipulating, running numerical simulations; exporting SBML models; and converting SBML models in to other formats, such as XPP, HTML and FORTRAN. By accessing the full breadth of Mathematica functionality, MathSBML is fully extensible to SBML models of any size or complexity.

AVAILABILITY

Open Source (LGPL) at http://www.sbml.org and http://www.sf.net/projects/sbml

Conservation analysis in biochemical networks: computational issues for software writers

Large scale genomic studies are generating significant amounts of data on the structure of cellular networks. This is in contrast to kinetic data, which is frequently absent, unreliable or fragmentary. There is, therefore, a desire by many in the community to investigate the potential rewards of analyzing the more readily available topological data. This brief review is concerned with a particular property of biological networks, namely structural conservations (e.g. moiety conserved cycles). There has been much discussion in the literature on these cycles but a review on the computational issues related to conserved cycles has been missing. This review is concerned with the detection and characterization of conservation relations in arbitrary networks and related issues, which impinge on simulation simulation software writers. This review will not address flux balance constraints or small-world type analyses in any significant detail.

The next step in systems biology: simulating the temporospatial dynamics of molecular network

As a result of the time- and context-dependency of gene expression, gene regulatory and signaling pathways undergo dynamic changes during development. Creating a model of the dynamics of molecular interaction networks offers enormous potential for understanding how a genome orchestrates the developmental processes of an organism. The dynamic nature of pathway topology calls for new modeling strategies that can capture transient molecular links at the runtime. The aim of this paper is to present a brief and informative, but not all-inclusive, viewpoint on the computational aspects of modeling and simulation of a non-static molecular network.

Molecular interaction maps--a diagrammatic graphical language for bioregulatory networks

Molecular interaction maps (MIMs) use a clear, accurate, and versatile graphical language to depict complex biological processes. Here, we discuss the main features of the MIM language and its potential uses. MIMs can be used as database resources and simulation guides, and can serve to generate new hypotheses regarding the roles of specific molecules in the bioregulatory networks that control progression through the cell cycle, differentiation, and cell death.

Simulating cellular dynamics through a coupled transcription, translation, metabolic model

In order to predict cell behavior in response to changes in its surroundings or to modifications of its genetic code, the dynamics of a cell are modeled using equations of metabolism, transport, transcription and translation implemented in the Karyote software. Our methodology accounts for the organelles of eukaryotes and the specialized zones in prokaryotes by dividing the volume of the cell into discrete compartments. Each compartment exchanges mass with others either through membrane transport or with a time delay effect associated with molecular migration. Metabolic and macromolecular reactions take place in user-specified compartments. Coupling among processes are accounted for and multiple scale techniques allow for the computation of processes that occur on a wide range of time scales. Our model is implemented to simulate the evolution of concentrations for a user-specifiable set of molecules and reactions that participate in cellular activity. The underlying equations integrate metabolic, transcription and translation reaction networks and provide a framework for simulating whole cells given a user-specified set of reactions. A rate equation formulation is used to simulate transcription from an input DNA sequence while the resulting mRNA is used via ribosome-mediated polymerization kinetics to accomplish translation. Feedback associated with the creation of species necessary for metabolism by the mRNA and protein synthesis modifies the rates of production of factors (e.g. nucleotides and amino acids) that affect the dynamics of transcription and translation. The concentrations of predicted proteins are compared with time series or steady state experiments. The expression and sequence of the predicted proteins are compared with experimental data via the construction of synthetic tryptic digests and associated mass spectra. We present the mathematical model showing the coupling of transcription, translation and metabolism in Karyote and illustrate some of its unique characteristics.

Developing complex signaling models using GENESIS/Kinetikit

The development of biologically realistic models of signaling pathways is a demanding process, involving computational challenges as well as those arising from the complexity of detailed pathway models. We have developed the General Neural Simulation System (GENESIS) and Kinetikit (GENESIS/Kinetikit), a graphical simulation environment for modeling biochemical signaling pathways using deterministic and stochastic methods. A library of models of several common signaling pathways complements the software. This combination of numerical computation engines, graphical modeling tools, and library of models is designed to build on the cumulative development of models and techniques from many sources. The complete simulation environment and demonstration models are available from (http://stke.sciencemag.org/cgi/content/full/sigtrans;2004/219/pl4/DC1; also at http://www.ncbs.res.in/~bhalla/kkit/download.html). The associated library of signaling pathways is based on published experimental and simulation studies and is curated to ensure that the simulation outcomes match published results. Models in the library are maintained in a database (http://doqcs.ncbs.res.in). Individual pathway models can be combined to build complex signaling network simulations. The overall goal of this process is to attain sufficient biological realism in models to directly compare their outcomes with experiments and to improve our understanding of complex signaling.

Towards a systems biology understanding of human health: Interplay between genotype, environment and nutrition

Sequencing of the human genome has opened the door to the most exciting new era for the holistic system description of human health. It is now possible to study the underlying mechanisms of human health in relation to diet and other environmental factors such as drugs and toxic pollutants. Technological advances make it feasible to envisage that in the future personalized drug treatment and dietary advice and possibly tailored food products can be used for promoting optimal health on an individual basis, in relation to genotype and lifestyle. Life-Science research has in the past very much focused on diseases and how to reestablish human health after illness. Today, the role of food and nutrition in human health and especially prevention of illness is gaining recognition. Diseases of modern civilization, such as diabetes, heart disease and cancer have been shown to be effected by dietary patterns. The risk of disease is often associated with genetic polymorphisms, but the effect is dependent on dietary intake and nutritional status. To understand the link between diet and health, nutritional-research must cover a broad range of areas, from the molecular level to whole body studies. Therefore it provides an excellent example of integrative biology requiring a systems biology approach. The current state and implications of systems biology in the understanding of human health are reviewed. It becomes clear that a complete mechanistic description of the human organism is not yet possible. However, recent advances in systems biology provide a trajectory for future research in order to improve health of individuals and populations. Disease prevention through personalized nutrition will become more important as the obvious avenue of research in life sciences and more focus will need to be put upon those natural ways of disease prevention. In particular, the new discipline of nutrigenomics, which investigates how nutrients interact with humans, taking predetermined genetic factors into account, will mediate new insights into human health that will finally have significant positive impact on our quality of life.

Kinetic simulation of signal transduction system in hippocampal long-term potentiation with dynamic modeling of protein phosphatase 2A

We modeled and analyzed a signal transduction system of long-term potentiation (LTP) in hippocampal post-synapse. Bhalla and Iyengar [Science 283(1999) 381] have developed a hippocampal LTP model. In the conventional model, the concentration of protein phosphatase 2A (PP2A) was fixed. However, it was reported that dynamic inactivation of PP2A was essential for LTP [J. Neurochem. 74 (2000) 807]. We introduced a dynamic modeling of PP2A; inactivation (phosphorylation) of PP2A by calcium/calmodulin-dependent protein kinase II (CaMKII) in the presence of calcium/calmodulin, self-activation (autodephosphorylation) of PP2A, and inactivation (dephosphorylation) of CaMKII by PP2A. This model includes complex feedback loops; both CaMKII and PP2A are autoactivated, while they inactivate each other. Moreover, we proposed an analysis strategy for model validation by applying the results of sensitivity analysis. In our system, calcineurin (CaN) played an essential role, rather than the activation of protein kinase C (PKC) as documented in the conventional model. From results of the analysis of our model, we found the following robustness as characteristics of bistability in our model: (1). PP2A reactions against calcium ion (Ca(2+)) perturbation; (2). PP2A inactivation against PP2A increase; (3). protein phosphatase 1 (PP1) activation against PF2A increase; and (4). PP2A reactions against PP2A initial concentration. These properties facilitated LTP induction in our system. We showed that another mechanism could introduce bistable behavior by adding dynamic reactions of PP2A.