Tutorials

Tutorials will be scheduled on Monday, October 1, prior to the main conference that runs from October 2 – 4, 2007. There are 18 tutorials, nine in the morning and nine in the afternoon. You may register for up to two tutorials, one in the morning and one in the afternoon. Each tutorial is approximately 3 hrs in length.

AM1. What is Systems Biology? A Survey of Systems Biology Research

Tau-Mu Yi and other members of the Center for Complex Biological Systems, University of California, Irvine, CA, USA

What is systems biology? This question has tormented the systems biology community; there are as many definitions as practitioners. Is there substance behind the hype? In this tutorial, we will attempt to survey the field using classic papers as a starting point. We will provide specific examples of the aims, approaches, methods, results, and conclusions of a cross-section of systems biology research. An emphasis will be placed on furnishing a multi-disciplinary perspective on important concepts such as spatial dynamics, regulatory networks, and robust behaviors. The goal is to lower the activation barrier for doing systems biology research for both novices and experts by highlighting the thought process rather than specific techniques.

The tutorial is expected to last 3 hours. The tutorial will cover 5 basic topics: Dynamics, Variation, Control, Networks, and Design. Each topic will revolve around one seminal paper in the area. The tutorial will dissect the papers in the context of related research.

The materials for this tutorial will be derived from the new curriculum being developed for the Mathematical, Computational, and Systems Biology (MCSB) graduate program in systems biology at UCI, http://mcsb.bio.uci.edu/index.html. In particular, we will make use of a sourcebook being compiled for a "Critical Thinking in Systems Biology" course.

The targeted audience are newcomers to the systems biology field. There are no prerequisites.

The expected outcomes and goals are that the attendees will be better prepared to propose, formulate and carry out systems biology research in the future. A secondary goal is that the tutorial can stimulate new ideas on teaching systems biology.

AM2. Mathematical Tools for the Analysis of Biochemical Network Dynamics

German A. Enciso, Mathematical Biosciences Institute and Eduardo D. Sontag, Rutgers University

An important goal in quantitative biology is to look for new analytical approaches to study gene and protein regulatory networks, as new measurements are allowing for an increasing complexity and accuracy of biochemical models. The proposed tutorial will give a general introduction to some of these methods, with an emphasis on determining specific dynamical properties such as global convergence towards an equilibrium, periodic oscillations, etc. We present various general mathematical results, with applications to concrete examples from the molecular biology literature.

Eduardo Sontag, David Angeli and collaborators have recently proposed in a series of papers to use the established theory of so-called monotone dynamical systems, to prove the global asymptotic stability of certain systems which are not themselves monotone. Similar ideas have led to the study of oscillations in cyclic time delay systems. These results have been applied to various models ranging from the lac operon to MAP kinase cascades, testosterone dynamics, and the somitogenesis oscillator. These ideas constitute the material for two of the lectures.

In his talk, Patrick de Leenheer will describe the concept of persistence: provided that every molecular species is present at the start of a biochemical reaction, no species will tend to be eliminated in the course of the reaction. He will provide checkable conditions for the persistence of various reaction networks, including models of cell signaling pathways.

Gheorghe Craciun will speak about the species-reaction (SR) graph of a chemical reaction system, and how it gives immediate information about the network's capacity for multiple steady states.

AM3. Stochastic Gene Expression in Systems Biology

Mustafa Khammash and Brian Munsky, University of California, Santa Barbara, CA, USA

The cellular environment is abuzz with noise. A key source of this "intrinsic" noise is the randomness that characterizes the motion of cellular constituents at the molecular level. Cellular noise not only results in random fluctuations (over time) within individual cells, but it is also a source of phenotypic variability among clonal cellular populations. In some instances fluctuations are suppressed downstream through an intricate dynamical networks that acts as noise filters. Yet in other important instances, noise induced fluctuations are exploited to the cell's advantage. Researchers are just now beginning to understand that the richness of stochastic phenomena in biology depends directly upon the interactions of dynamics and noise and upon the mechanisms through which these interactions occur.

Tutorial Description: In this tutorial we review a number of approaches for the analysis of stochastic fluctuation in gene expression. We will explore: a) analytical and computational methods for the analysis of stochasticity in living cells; and b) examples of gene regulatory networks that suppress or exploit noise, including discussion of landmark papers that report measurements of stochasticity and its impact on biological function.

Specific topics include: Introduction to stochastic gene expression; Deterministic vs. stochastic models; The stochastic chemical kinetics framework; A rigorous derivation of the chemical master equation. Moment computations; Linear vs. nonlinear propensities; Linear noise approximations; Monte Carlo simulations; Gillespie's Stochastic Simulation Algorithm; Variants of the SSA; Direct methods for the solution of the Chemical Master Equation; Finite State Projections; Moment Closure methods; Intrinsic and extrinsic noise in gene expression. Propagation of noise in cell networks; Noise suppression in cells; The role of feedback; How cells exploit noise; Noise focusing; Coherence resonance; Competence in B. Subtilis; Bimodality and stochastic switches; The pap pili switch.

AM4. Computational analyses across the BioCyc collection of Pathway/ Genome Databases

Peter Karp and Suzanne Paley, SRI International, Menlo Park, CA, USA

BioCyc is a collection of 260 pathway/genome databases for most organisms whose genomes have been completely sequenced. It is a large and comprehensive resource for systems biology research. We expect that many bioinformatics and computational biology researchers will be interested in computing with BioCyc to address global biological questions, such as studying the phylogenetic distribution and evolution of metabolic pathways. The goal of this tutorial will be to provide researchers with the information they need to perform global analyses of BioCyc. The tutorial will cover the methodologies used to create BioCyc, a description of the complex database schema and ontologies that underlay BioCyc, and descriptions of the APIs that are available to query BioCyc. The tutorial will also present the Pathway Tools semantic inference layer, which is a library of commonly used queries that we have encoded to save researchers time. We will also consider common stumbling blocks and misconceptions that can lead to misinterpretations of the data.

Expected outcomes and goals: Students will learn how to perform computational analyses across the large BioCyc collection of Pathway/ Genome Databases.

Prerequesites: Basic familiarity with programming and databases and basic familiarity required with concepts in biology and metabolic

AM5. Advanced Model Analysis with COPASI

Pedro Mendes, University of Manchester, UK
Stefan Hoop, Virginia Bioinformatics Institute, USA
Sven Sahle, University of Heidelberg, Germany

We will explain how to utilize parameter scan, optimization, and parameter estimation to understand and improve models using the COPASI software. The tutorial will include a short introduction to the optimization problem.

COPASI (Complex Pathway Simulator) is a software application for simulation and analysis of biochemical networks. It is developed jointly by the groups of Pedro Mendes (Virginia Bioinformatics Institute, USA, and University of Manchester, UK) and Ursula Kummer (University of Heidelberg, Germany), and is freely available for academic use.

COPASI's current features include stochastic and deterministic time course simulation, steady state analysis (including stability), metabolic control analysis, elementary mode analysis, mass conservation analysis, import and export of SBMLlevel 2, optimization, parameter scanning and parameter estimation. It runs on MS Windows, Linux, OS X, and Solaris SPARC.

Participants are strongly encouraged to bring their own computers.

Target Audience: This tutorial is primarily aimed at experimentalists who are newcomers to the computational side of systems biology or experienced modelers who want to explore advanced parameter estimation features of COPASI.

AM6. Systems Biology Toolbox for MATLAB Cancelled

AM7. Drawing, annotating and analyzing biological pathways with Edinburgh Pathway Editor

Stuart L. Moodie and Anatoly Sorokin University of Edinburgh, UK

Summary. The Computational Systems Biology Group at the University of Edinburgh, UK, would like to present a tutorial focusing on the Edinburgh Pathway Editor (EPE). This would cover how to use EPE to reconstruct, annotate and analyze different types of biological pathway - including signal transduction and metabolic pathways. The tutorial would be very much an interactive exploration of the software covering what we believe to be an increasingly important topic in Systems Biology.

Description. The tutorial will take participants through the steps of reconstructing, annotating and analyzing biological pathways using graphical notations within EPE. Participants will work through specific examples and will learn how to draw pathway maps using the notations supported by the software including the Edinburgh Pathway Notation (EPN), Kitano notation, KEGG and EMP. We will also provide instruction on how to draw metabolic and signaling pathways and provide advice on how EPE's annotation facilities can help when reconstructing such pathways from the literature and online resources. New in EPE are its auto-layout facilities. We will show users how they can create and automatically layout new maps by importing networks from external sources; we will also show how participants can use auto-layout and EPE's drawing facilities to obtain an aesthetically pleasing layout (not easy when drawing a large map by hand). Finally, the participants will learn how to export their maps from EPE to a variety of file formats, which will allow them to analyze their pathways using other software.

Tutorial Outline: Introduction and installation of EPE; Overview of graphical notations used in Systems Biology; Reconstructing a metabolic network in EPE; Reconstructing a signal transduction pathway; Adding annotation and web links to you map; Using the auto-layout functionality in EPE; Exporting maps from EPE for further analysis.

Background. Edinburgh Pathway Editor (EPE) is a freely available visual editor designed to support the visualization, annotation and analysis of a wide variety of biological networks, including metabolic, genetic and signal transduction pathways. It has a metadata driven architecture, which makes it very flexible in drawing, storing, presenting and exporting information related to the network of interest. This architecture enables it to support a variety of graphical notations for signal transduction and metabolic pathways.

Organization. The tutorial will be interactive and participants will be strongly encouraged to bring a laptop to use during the tutorial. We will make all tutorial materials and copies of the software available on our website in advance of the conference. All though not essential, internet access for users would be useful.

AM8. The Systems Biology Workbench

Frank Bergmann^1,2, Deepak Chandran¹, and Herbert M. Sauro ¹, ¹University of Washington, Seattle, WA, USA and ²Keck Graduate Institute, Claremont, CA, USA

The Systems Biology Workbench (SBW) is an extendable, open source software framework, connecting software applications written in a variety of programming languages. Software components provided with SBW assist in analyzing, creating, optimizing, simulating and visualizing computational models.

This tutorial aims to familiarize participants with the tools provided in the Systems Biology Workbench to aid them in many aspects of systems biology research. We will introduce modeling concepts, time course and steady state concepts, metabolic control analysis and bifurcation analysis at the example of our software tools, or 3rd party tools integrated in our framework.

Outline: Introduction; Modeling (Continuous, Stochastic, Jarnac, JDesigner); Hands-on Exercises (Oscillators and Bistable switches); Simulation and Visualization (Time course analysis; steady state analysis); Hands-on Exercises (Homeostasis, feed-forward networks, MCA); Analysis (Bifurcation Discovery Tool, Oscill8, Frequency Analysis, Stochastic Simulation); Hands-on Exercise (Noise in reaction networks)

The target audience for this introductory tutorial will be research scientists interested in modeling as well as simulation and analysis of computational models. Participation in the tutorial does not require prior experience in modeling/ simulation or skills in computer programming. Tutorial material along with a software release will be made available on http://sys-bio.org.

AM9. PySCeS: the Python Simulator for Cellular Systems

Johann M. Rohwer and Brett G. Olivier, Stellenboesch University, Stellenboesch, South Africa

Computer modeling has become an integral tool in the analysis and understanding of the reaction net- works that underlie cellular processes. As such, numerous software packages have been developed for simulating and analyzing such networks (see e.g. http://sbml.org), each with its own advantages and disadvantages. The need for a flexible, customizable and extensible simulation system prompted us to develop PySCeS, the Python Simulator for Cellular Systems [http://pysces.sourceforge.net]. PySCeS is open-source, multi-platform software. It is built on the programming language Python and makes use of the SciPy (http://www.scipy.org) library of scientific tools for Python. PySCeS sup- ports the following types of analysis: structural analysis including calculation of elementary modes, time-course simulation, solving for steady-state, control analysis, stability analysis and eigenvalue determination, data output in LATEX and HTML format, and model import and export in Systems Biology Markup Language (SBML). Simulation results can be graphed with the interface to the Mat- plotlib (http://matplotlib.sourceforge.net) plotting library. One of PySCeS's particular strengths is its modular design that allows it to take full advantage of Python's ability to interface with numerical routines implemented in Fortran and C. Such routines can then be directly accessed from within PySCeS.

The purpose of this tutorial is to give the participants a general introduction to modeling with PySCeS, as well as to present an overview of the program's features. "Snapshot previews" of the advanced features will be included to demonstrate the range of the program's capabilities. The following topics will be covered: Defining a model - PySCeS input file syntax, SBML import; Basic analysis - Structural analysis, time-course, steady state; Data visualization - Model visualization, graphical output, Web reports, data output for analysis in a spreadsheet; Advanced model analysis - Control analysis, one- and multi-dimensional parameter scans; Software extension - Your own integration / solver / optimization algorithm; Parallel computing.

Ideally, the tutorial will be hands-on. Participants can use their own laptops and software will be supplied (either by download or on a CD). PySCeS supports Windows, Linux and Mac OS X.

PM1. New Mathematical Methods for Systems Biology Sold Out

Eric Mjolsness, University of California, Irvine, CA, USA

Expectations and ambitions for the future of computational systems biology are ever growing, but several significant problems of applied mathematics and modeling stand in the way. These problems include the relations between stochastic and deterministic models and simulation algorithms, adequate models of molecular complexes, the role of spatial inhomogeneity at subcellular and multicellular scales, modeling biological graph structure and dynamics, inference from heterogeneous data sets, and the reuse and integration of modeling techniques across spatial scales from molecular to developmental and ecological.

Fortunately, there are relevant branches of applied mathematics that have been underexploited in attacking these problems, and it's not too hard to understand their foundations. I suggest that the basic mathematical toolkit for systems biology will come to include not only such staples as differential equation and graphical probabilistic models, but also operator algebras, context sensitive grammars, stochastic field theory of both particle-like and extended objects, partition functions, multiscale modeling, aspects of algebraic geometry, and dynamical systems defined on static and dynamic graphs. I will explain why, what, and how, and give examples from many spatial and temporal scales: bacterial metabolism, eukaryotic transcriptional regulation and signal transduction, developmental biology of plants including phyllotaxis, and population biology.

Tentative outline: Part I, Elementary methods: Biological problem formulation, probabilistic models and information, differential equation dynamics, graph operations, formalization tools; applications to cellular systems. Part II, Advanced methods: Operator algebra dynamics, indexed and parameterized reaction schemata, equilibrium and nonequilibrium statistical mechanics, inference methods, geometry, homology; applications to developmental systems. Notes (150pp).

PM2. Genetic Algorithms and their Application to the Artificial Evolution of Genetic Regulatory Networks

Katja Wegner, Johannes Knabe, Mark Robinson, and Maria Schilstra, University of Hertfordshire, Hatfield, UK

Purpose: To provide an audience who are unfamiliar with computational evolutionary techniques with a general introduction to genetic algorithms (GAs); to illustrate the applicability and limitations of GAs using artificial evolution of genetic regulatory networks as an example. Background

Background: Genetic algorithms are a class of evolutionary algorithms that are inspired by the process of natural selection and evolution in biological systems. GAs are generally employed in attempts to solve optimization problems where the fitness landscape is complex. The application of a GA to a specific problem such as the construction of GRNs requires tailored specification of 1) the "genome" (a representation of the solution domain that allows modification by evolutionary operators, such as mutation and recombination), 2) a "fitness function" to evaluate candidate solutions, and 3) appropriate "reproduction", "selection", and termination strategies.

The reconstruction of genetic regulatory networks (GRNs, networks of interactions between genes and gene products) is a necessary precursor for gaining a functional understanding of the topological and dynamical properties of these networks. GAs may be of considerable use to reveal potential interactions and parameters in partially reconstructed GRNs. Alternatively, evolving artificial GRNs de novo can give new insights into the constraints that real GRNs are subject to.

Tutorial outline.: In the first part of the tutorial we will outline the fundamentals of GAs and their application to the solution of optimization problems, and the basic concepts that underlie their functioning. We will also briefly discuss current ideas about the function, structure, and dynamic behavior of GRNs, and indicate how these features are typically modeled and simulated. Emphasizing the fact that different systems require different approaches, we will then use our own work on the evolution of GRN-like control networks as a concrete example to demonstrate how to construct a mutable "genome" and an initial "populations" of candidate solutions, how to assess the "fitness"of each individual, and how to apply "selection pressure" and "evolve" the population towards one in which all individuals exhibit a pre-defined target behavior.

PM3. Rule-Based Kinetic Modeling of Signal Transduction Networks

Michael L. Blinov¹, James R. Faeder², William S. Hlavacek²; ¹University of Connecticut Health Center, Farmington, CN, USA; ²Los Alamos National Laboratory, Los Alamos, NM, USA

Cell signaling, the process by which cells sense and respond to their environment, involves a large number of proteins and other biomolecules whose interactions define a vast response network. A key feature of these systems is that the molecules involved have a modular structure that allows each molecular component of the network to interact with a large number of other elements. Modeling the dynamics of such complex systems poses a number of challenges, but is critical for developing a mechanistic understanding of biological signal transduction and the ultimate goal of controlling pathological responses to cure and prevent disease. In this tutorial, we will describe how to develop and simulate kinetic models of signaling networks using a simple yet powerful language (BioNetGen language, BNGL) and software (BioNetGen) we have developed. BNGL allows explicit representation of the individual elements that mediate the interactions among proteins and other signaling molecules. For example, molecules are represented as structured objects in which the functional elements are sites that may bind to other sites of the same or different molecules and which may have an associated internal state that represents either conformation or covalent modification. The model is built by defining rules that govern how molecules interact to form complexes, modify internal states, and degrade or produce new molecules. The application of rules to a seed set of molecules is used to generate a reaction network, freeing the user from the intense bookkeeping that would be required to enumerate such a network by hand and greatly reducing the barrier to exploring how alternate formulation of the rules would affect model behavior. We will describe a number of options for simulating network kinetics, including ODE's and kinetic Monte Carlo using the popular Gillespie algorithm.

We will demonstrate how exporting models in the Systems Biology Markup Language (SBML) provides compatibility with a large number of additional simulation tools and methods. We will show how to define macroscopic variables, which represent quantities that can be directly compared with experimental data, such as Western blots and coimmunoprecipitation. The tutorial will provide hands-on experience on how to model and simulate portions of signaling pathways (using the web-version of BioNetGen), describing several published models and discussing how they can be extended in the future. We will discuss how the rule-based description could be used as a way to represent knowledge about the interactions present in signaling networks and how it could provide the basis for a collaborative framework aimed at developing comprehensive models of signaling pathways.

PM4. PANTHER Pathway Curation System: An infrastructure for community curation and contribution of biological network knowledge

Huaiyu Mi, Anish Kejariwal, Nan Guo, and Paul Thomas, SRI International, Menlo Park, CA, USA

The PANTHER (Protein ANalysis THrough Evolutionary Relationships) Classification System is freely available at http://www.pantherdb.org. It was designed to model evolutionary sequence-function relationships on a large scale. Its core is a library of a large collection of protein families that have been subdivided into functionally related subfamilies, using human expertise. These subfamilies model the divergence of specific functions within protein families, allowing more accurate association with function, as well as inference of amino acids important for functional specificity. Hidden Markov models (HMMs) are built for each family and subfamily for classifying additional protein sequences.

PANTHER pathway (http://www.pantherdb.org/pathway) is one of the modules of the PANTHER System. There are 3 major characteristics of the system. First, the pathways were generated using the emerging SBML standard using the CellDesigner pathway network editing software. As a result, detailed molecular events of biochemical reactions are captured from the diagrams, and stored in file in SBML format, which keeps consistency between the data and the diagrams. Second, all pathway diagrams can be viewed in a simplified relationship diagram similar to those in most scientific papers, which provides a user-friendly user interface for biologists. Third, all pathway components are directly linked to protein sequences from the PANTHER library through manual curation, connecting pathways to molecular phylogenetic and genomic data. Therefore, various PANTHER web tools, including the protein classification tool and gene expression tool, are linked to pathways. As a result, it becomes a powerful system for users to predict protein function, protein relationships, and analyze experimental results. This demo will cover the curation infrastructure for generating pathways.

The PANTHER Pathway curation is available on the web (http://curation.pantherdb.org). It is composed of 2 phases. The first phase is to generate pathway diagram and ontology. During this phase, a biologist curator draws a pathway diagram using CellDesigner. Literature references must be provided for the pathway. The CellDesigner file that is created during this step adheres to SBML format. We have developed a parser that reads the SBML and uses the information to create a pathway ontology, which is then stored in the PANTHER Pathway curation database, implemented in Oracle. The second phase is to link pathway to protein sequences in PANTHER protein library. During this phase of curation, the curator works with a direct web interface to the curation database. The interface displays each of the ontology classes (terms) that correspond to a protein, mRNA or gene. The curator associates each term with individual protein sequences that are instances of the class as described above. Upon completion of curation, all pathways can be reviewed and published on the PANTHER Pathway website with proper author acknowledgements. Therefore, all the curated pathways will be linked to various tools for protein classification, gene expression data analysis, and SNP data analysis.

PM5. CellDesigner 4.0: A Process Diagram Editor for Gene-Regulatory and Biochemical Networks

Akira Funahashi^1,2, Akiya Jouraku^1,2, Yukiko Matsuoka¹, Norihiro Kikuchi³ and Hiroaki Kitano^{1. 1}The Systems Biology Institute, Japan; ²Keio University, Japan; ³Mitsui Knowledge Industry Co., Ltd., Japan

CellDesigner is software for modeling and simulation of biochemical and gene regulatory networks, originally developed by the Systems Biology Institute in Japan. While CellDesigner itself is a sophisticated structured diagram editor, it enables users to directly integrate various tools, such as built-in SBML ODE Solver and SBW-powered simulation/analysis modules. CellDesigner runs on various platforms such as Windows, MacOS X and Linux, and is freely available from our website at http://celldesigner.org. In this course, we will explain how CellDesigner can be used from both modeling and software development perspectives. The first topic will feature network modeling using CellDesigner, and will show how she/he could build a model from scratch, and examine simulations. The second topic will feature plugin development of CellDesigner, which allows users to manipulate network diagram in many ways (for example changing the color/size of node, reflecting experimental data etc.). This tutorial will cover both modeling and software development topics, thus both CellDesigner users and software developers are encouraged to join. Bringing your notebook PC is highly recommended.

PM6. Inverse Methodologies for Systems Biology: SOSlib, MathSBML and Matlab extensions

James Lu¹, Stefan Müller¹, Christoph Flamm² and Rainer Machne²; ¹Radon Institute for Computational and Applied Mathematics, Linz, Austria; ²University of Vienna, Vienna, Austria

In this tutorial we introduce a variety of advanced inverse methods for numerical analysis of SBMLencoded biochemical models using experimental data. Inverse problems, e.g. parameter identification or probing for possibility of multistability and/or oscillations for a given model, are typically ill-posed, i.e. the solution may be non-unique and unstable with respect to noise in experimental data. The tutorial is organized in three parts, 1 hour each. Links to prerequisite software and background material are available at: http://www.tbi.univie.ac.at/wiki/index.php/ICSB07 tutorial.

Forward Analysis: The SBML ODE Solver Library (SOSlib) is a generic ISO C/C++ programming library for the numerical analysis of SBML models, with bindings for e.g. Java and Perl. The tutorial will introduce basic data structures and interfaces of SOSlib on code examples in Perl, Java and C. Only a couple of lines of code are required to implement applications ranging from multi-scale modeling with communicating integrator instances, efficient parameter and initial condition scans to sensitivity analysis. Inverse Analysis: SOSlib functionality can be employed for a combination of local and global search strategies to identify unknown kinetic parameters in an SBML model from experimental data. To stabilize the solutions w.r.t. data noise, various so-called regularization techniques are applied. The identification problem is then formulated as a penalized optimization problem and is solved using forward and adjoint capabilities of SOSlib within the interior point optimizer IpOpt, and scatter search as a globalization strategy.

Inverse Dynamical Analysis: To probe the possibility of a biological model to undergo saddle-node or Hopf bifurcations the Inverse Eigenvalue Analyzer, an add-on to the MathSBML Mathematica package, attempts to place the minimal eigenvalues of the system onto the origin or the imaginary axis, respectively. To infer which core regulation mechanisms underlie the bifurcation points, an inverse bifurcation analysis is carried out by the Matlab add-on Inverse Bifurcation Toolbox. Here regularization is used to promote the sparsity of the solution, i.e. to identify minimal sets of parameters that need to be modified to realize a given bifurcation pattern. A hierarchical algorithm allows to identify several alternative minimal parameter sets to achieve the sought-for dynamics.

PM7. Computational Cell Biology with VCell

Ion I. Moraru and James C. Schalff, University of Connecticut Health Center, Farmington, CT, USA.

The Virtual Cell (VCell; http://vcell.org) is a unique software environment for computational cell biological research developed at the Richard D. Berlin Center for Cell Analysis and Modeling (CCAM) at the University of Connecticut Health Center. CCAM is a NIH Technology Center for Networks and Pathways and a NIH-designated National Research Resource. The center integrates new microscope technologies for making quantitative in vivo live cell measurements with new physical formulations and computational tools that will produce spatially realistic quantitative models of intracellular dynamics. The latter are being made available for the use of researchers worldwide through their gradual integration into the public, web-accessible, VCell framework.

VCell has been continuously and rapidly growing in capabilities and complexity over the past several years. To date, more than 1,000 independent users worldwide have created and run simulations with VCell. Since 2001 we have regularly organized tutorials, workshops, or other forms of public presentations at various meetings (such as Biophysical Society, ICSB, ASCB, Computational Cell Biology, etc). The focus was both on general issues in quantitative modeling in cell biology and on introducing new features of VCell through demonstrations and hands-on interactions.

The proposed tutorial will showcase some of the many new capabilities of VCell (parallel solvers, stochastic solvers, flow/advection, grid computing) as well as discuss two other major developments that are in progress: the transition of VCell to Open Source, and the introduction of specialized, stand-alone VCell applications, external tools, and a new plug-in architecture. Consequently, this year's ICSB tutorial will target researchers and modelers, as well as computer scientists and developers.

The tutorial will include:

• two short talks: (i) presenting the concepts and abstractions underlying the use of the Virtual Cell for building models and running simulation, including a discussion of practical and theoretical issues in spatial modeling and biological networks, and (ii) presenting the new architecture and design of the VCell software framework, including how to develop stand-alone applications and external plugins

• a demonstration session: (i) a typical sequence of building a simple model, creating an application, running simulations, and viewing and exporting results, with the web-based version of VCell, (ii) a presentation of several more complex models present in the public database, illustrating some of the advanced features of the software, (iii) a presentation of specialized stand-alone VCell applications, such as the Virtual Microscopy tool and the Virtual FRAP tool

• a hands-on session: (i) attendees that have laptops can run the web-based and standalone versions of VCell, browse existing public models or create their own, (ii) open discussion of features, capabilities, and any other technical details (including source-code and API features) with members of our team

PM8. Formal description and visual modeling of complex biological systems using BioUML workbench and BioUML Network Edition

Fedor Kolpakov, Institute of Systems Biology, Novosibirsk, Russia

The purpose of this tutorial is to demonstrate and teach how new possibilities of BioUML Workbench and BioUML Network Edition can be used for formal description and simulation of complex biological systems. Several examples will be considered: NF-kappa B pathway; cell cycle; apoptosis; arterial hypertension, including simulation of blood flow.

BioUML, Biological Universal Modeling Language, is an open source extensible Java workbench for systems biology. Its core is a meta model that provides an abstract layer for comprehensive formal description of wide range of biological and other complex systems. Content of databases on biological pathways, SBML and CellML models, as well as databases in BioPAX format can be expressed in terms of the meta model and used by BioUML workbench.

New version BioUML Workbench and BioUML Network Edition (to be released August 2007) provide new possibilities for formal description and visual modeling of complex biological systems. We believe some of these possibilities to be revolutionary. Below list of main new features of BioUML Workbench and BioUML Network Edition: full text search using Lucene search engine; new graph search engine and improved graph layout algorithms; possibility of seamless integration of external databases into user's database. For example Ensembl database can be used as gene catalogue in the user's database.; import/export data in BioPAX format; graphic notation editor - allows user to create their own graphic notation and corresponding diagram types; arterial tree diagram type - used for simulation hemodynamics (blood flow); composite diagram - a mechanism to combine several models into one bigger model using the same mathematical formalism; agent based diagram - a mechanism to combine several models into one bigger model using different mathematical formalism. For example one model is described by system of PDE and other model is described by system of ODE. Additionally models composition is dynamical - models (agent) can appear, disappear (die) and move in space; BioUML server - provides high level protocol for BioUML workbench for data access on server side, as well as for data search (full text search, graph search); publishing BioUML data using BeanExplorer Enterprise Edition technology; ru.biosoft.bsa plug-in - BioSequence Analyses - provides visualization and analyses of biological (mainly nucleotide) sequences. This is updated version of library that is core of TRANSPLORER tool; org.openscience.cdk plug-in - Chemical Development Kit - allows to visualize structural formulas for chemical substances on diagrams.

During the tutorial attendees will know and obtain hands-on experience on how to:use existing databases on biological pathways: Reactome, TRANSPATH, KEGG/Pathways, GeneOntolgy, BMOND, GeneNet and some other) from BioUML workbench; BMOND, Cyclonet and LipidNet databases as examples of databases created using BioUML technology ; create their own database; specify external databases (Ensembl, GeneOntolgy, Reactome, TRANSPATH, KEGG and some other) as external catalogues (modules) for user's database; database search ; graph layout; create several diagram types to describe biological system on several semantic levels:; create own graphic notation using graphic notations editor; work with library of chemical kinetic laws; import/export models to SBML format; import/export databases in BioPAX format; use Java simulation engine; use MATLAB simulation engine; write and visualize simulation results as plots; import microarray data; bind microarray data with diagrams; bind results of microarray data analyses (for example up/down regulated genes) ; analyze gene regulatory regions, including search for transcription factor binding sites and composite elements; search results visualization; use JavaScript to automate simulation and data analysis; use BioUML workbench in console mode; publish user database using BioUML Network Edition

PM9. Accessing and annotating kinetic data for quantitative modeling: The SABIO-RK database

Martin Golebiewski and Ulrike Wittig, EML Research, Heidelberg, Germany

Biochemical model simulations need reliable quantitative data describing the dynamics of biological networks. To provide such data, we have developed SABIO-RK, a database system offering information about biochemical reactions and their corresponding kinetics. It describes participants and modifiers of the reactions, as well as measured kinetic data (including kinetic law equations) embedded in their experimental and environmental context.

SABIO-RK can be accessed in two different ways: through a web-based user interface to browse and search the data manually, and through web-services that can be automatically called by external tools, e.g. by integration tools of other databases or simulation programs. In both interfaces, reactions with their kinetic data can be exported in SBML (Systems Biology Mark-Up Language).

In this tutorial we will give an introduction into the usage of the SABIO-RK system. We will demonstrate how kinetic data can be found and retrieved for the set-up of quantitative biochemical models. A hands-on session will offer participants the opportunity to search the database in order to set-up a preliminary model of a biochemical reactions network that can then be exported in SBML format. Additionally, we will introduce our approach to standardize the data in SABIO-RK by annotating entities, using controlled vocabularies and bundling synonymic notations (e.g. for chemical compounds). Available web-service methods will be briefly introduced to provide details on how direct access to SABIO-RK can be integrated into modeling platforms or other databases.

Web Site: http://sabio.villa-bosch.de/SABIORK