SBML, BioSPICE Emerging Standards and Platforms for Systems Biology

1
SBML, BioSPICE - Emerging Standards and
Platforms for Systems Biology

• Eduardo Mendoza
• Dec 17, 2002
• Mathematics Department Sektion Physik
• University of the Philippines Ludwig-Maxim
  ilians-Universität

2
Topics

• What is Systems Biology?
• Why SBML?
• How can BioSPICE help?

3
1. Systems Biology just buzz for big bucks?

• Systembiologie
• alter Wein in neuen Schläuchen?
• (Laborjournal, 07-08/02)
• ...Not the first attempt at system-level
  understanding ..a recurrent theme in the
  scientific community
• (H. Kitano, ICSB 2000)

• BMBF Systeme des Lebens Project
• announced Dec 01, 50 m
• liver cell focus
• Initial awards for 15 m in Jan 03 to Uni
  Freiburg, Tübingen Rostock
• DARPA BioSPICE
• larger part of Bio-computing project started Fall
  01 60 m
• Vision provide bioscientists a standard,
  scalable, easy-to-use modeling and simulation
  environment

4
Systems Biology (1)

• Aims at systems-level understanding which
  requires a set of principles and methodologies
  that links the behaviors of molecules to systems
  characteristics and functions (H. Kitano,
  ICSB 2000)
• Modular Biology
• as advocated in the influential paper (Nature
  402, Dec 1999)

5
Lifes Complexity Pyramid (Oltvai-Barabasi,
Science 10/25/02)
6
Systems Biology (2)

• Need insight in 4 key areas
• Systems structures cf. above
• Systems dynamics eg sensitivity analysis,
  bifurcation analysis
• Control methods mechanisms for minimizing
  malfunction
• Design methods modify, construct biosystems with
  desired properties
• (Easy-to-use) Formula

First approximation (J. Schwaber, TJU, Nov 01)
Systems Biology Genomics Systems Engineering
Second approximation (MPI Magdeburg graphic,
2002) Systems Biology Biology Informatics
Systems Engineering
7
Optimal Formula Evolving Interdisciplinarity
Systems Engineering
Biology
Models of Biological Systems
Biosciences
Informatics
Biochemistry
Mathematics
Computational Sciences
Biophysics
Statistics
Systems Biology Biosciences Computational
Sciences Systems
Engineering ...
8
Systems Biology Research Cycle Vision
H. Kitano Science Mar 1 2002
9
Elementary modes in E.Coli metabolism

• Elementary (flux) mode analysis introduced by
  Schuster Hilgetag (1994)
• References
• J. Stelling, S. Klamt, K. Bettenbrock, S.
  Schuster E.D. Gilles Metabolic network
  structure determines key aspects of functionality
  regulation, Nature 420 (14/11/02)
• A. Cornish-Bowden, M.L. Cardenas Metabolic
  balance sheets, Nature 420 (14/11/02)
• S.Schuster, C. Hilgetag, J.H.Woods, D.A.Fell
  Reaction routes in biochemical systems Algebraic
  properties, validated calculation procedure and
  example from nucleotide metabolism Journal of
  Mathematical Biology 45 (2002)

10
Elementary (flux) mode example
11
Elementary modes mathematical definition and
properties

• Definitions
• stoichiometry matrix N
• index set of zero coordinates S(v)
• vectors in Ker N (null space) with sign
  restriction
• (flux) mode M
• elementary (flux) mode (also called direct
  reaction route)
• Concepts related to elementary modes direct
  mechanism (chemistry), minimal T-invariant
  (Petri nets)
• Algorithm for elementary modes is
  extension/adaptation of convex basis algorithm of
  Nozicka (1974)

12
Elementary modes in E.coli some results
analysis
Stelling et al Nature Nov 14 02
13
Understanding experiments(Cornish-Bowden,
Cardenas Nature Nov 14 2002)
14
Elementary modes in E.coli prediction of gene
expression

• Direct correlation between metabolic fluxes and
  transcriptome/proteome patterns not yet observed
• Concept of a reactions control-effective flux
  introduced to measure flexibility-efficiency
  trade-off (average flux weighted by mode
  efficiency)
• Ratio of control-effective fluxes ( theoretical
  transciption rates) for E.coli growth on
  different substrates compared with DNA microarray
  data

(Stelling et al Nature Nov 14 02)
15
Prediction details 1
16
Prediction details 2
17
Prediction details 3
18
Prediction vs. experiment
19
Elementary mode analysis prediction potential

• Examples
• (Cornish-Bowden, Cardenas Nature Nov 14 2002)
• Tolerable combination of mutations
• Genetic modifications enabling new properties
• Improvement of yield
• Robustness behavior

20
Network Motifs Building Blocks of Complex
Networks?
R. Milo et al, Science Oct 25 02
21
Functional units defining functional modules

• Criteria
• common physiological task eg specific catabolic
  pathways for individual carbohydrates
• common genetic units genes of all enzymes of an
  FU are organized in genetic units in a
  hierarchical structure (operons, regulons,
  modulons)
• common signal transduction network signal flow
  over the FU border (cross-talk) is small
  compared with information exchange within the
  unit

Kremling et al, Metabolic Engineering,2000-2001
22
2. Why SBML?

• Language for representation and exchange of
  biochemical network models
• Problems addressed (after Hucka et al 10/02)
• users often need to work with complementary
  resources from multiple tools gt manual
  re-encoding in each tool
• when simulators are no longer supported, encoded
  models become unusable
• Models published in peer-reviewed journals are
  not straightforward to examine and test as they
  use specific representation and environments.
  This also prevents a re-use strategy in building
  more complex models

23
SBML Evolution

• Software Platfoms for Systems Biology forum
  initiated April 2000 by ERATO Symbiotic Systems
  Project
• Principal Investigators H. Kitano (Keio
  U/Sony), J. Doyle (Caltech)
• Modeling/Simulation teams involved
• Berkeley Biospice (Arkin, UC Berkeley)
• Cellerator (Shapiro/Mjolsness, Caltech)
• DBsolve (Goryanin, Glaxo-Wellcome Research, UK)
• E-Cell (Tomita, Keio U)
• Gepasi (Mendes, Virginia Tech)
• Jarnac (Sauro, Caltech/KGI)
• StochSim (Morton-Firth/Bray, Cambridge U)
• Virtual Cell (Schaff, U Connecticut)
• ProMoT/DIVA (Ginkel, MPI Magdeburg)
• CellML (Hedley, U Auckland Physiome Sciences)

24
SBML Key Characteristics

• Based on XML ( further XML-based standards like
  MathML)
• Releases (called levels) community driven
  (sbml-discuss list)
• Key authors M. Hucka, A. Finney, H. Sauro
• Level 1 published Mar 01
• Level 2 to be published shortly
• Most tools mentioned already support SBML Level 1
• Convergence with CellML actively pursued
• Close affiliation with ERATO Systems Biology
  Workbench project ( www.sbw-sbml.org) at Caltech

25
What is XML (Extensible Mark-Up Language)?

• rapidly emerging IT industry standard for
  structured documents
• Developed by W3C 96/97 to overcome HTML
  limitations

XML W3C (96/97)
SGML (Standard Generalized Mark-Up Language) ISO
8879 (1986)
GML (IBM 70s)
HTML (T. Berners-Lee CERN (1991)
26
SBML 2 Model Definition
27
SBML 2 Examples Compartments
28
SBML 2Example for a Rule
gt Need Tools!
29
SBML 2 Hierarchy of Major Data Types
30
Current SBML 3 Plans (1)
31
Current SBML 3 Plans (2)
32
3. DARPA BioSPICE Project

• SPICE
• Originally Simulation Program for Integrated
  Circuit Evaluation
• DARPA Simulation Program for Intra-Cell
  Evaluation
• DARPA Bio-Computing
• Initiated Fall 01
• 2 Parts DNA Computing and BioSPICE (bigger)

• BioSPICE Principal Investigators
• Adam Arkin, UC Berkeley
• Roger Brent, The Molecular Sciences Institute,
  Berkeley
• John Byrne, University of Texas, Houston
• James Collins, Boston University
• John Doyle, Caltech
• Ron Fearing, UC Berkeley
• Tom Garvey, SRI International
• Megan Jewett, Brigham and Womens Hospital,
  Harvard
• David Kahaner, ATIP/ERATO
• James Liao, UCLA
• Bud Mishra, NYU Cold Spring Harbor Labs
• Arch Owen, BBN Technologies
• Harvey Rubin, U of Pennsylvania
• John Schwaber, Thomas Jefferson U
• John Tyson, Virginia Tech

33
BioSPICE Mission

• Provide bio-scientists a standard, scalable and
  easy-to-use modeling and simulation environment
  (as open source software) by
• Developing a model library (or kernel) of
  relevant biochemical processes and analytical
  tools
• Integrating this kernel into a simulation
  environment with links to relevant databases (eg
  genomic, protein interaction information)
• Including user interface and data visualization
  tools to facilitate use by experimenters
• Testing, refining and validating the models using
  a range of cell experiments

34
BioSPICE Architecture Requirements
35
BioSPICE based on OAA and SBML
36
BioSPICE Releases (every 6 months)

• 10 Validated Agents in BioSPICE 2.0 (just
  released)
• BioSketchpad
• GRASS (Gene Regulatory Analysis and Stochastic
  Simulation)
• GeneScreen
• Gillespie Chemical Simulation Module
• JigCell Model Builder
• Simpathica (Simulation of Pathways and
  Integrated Communications Analysis)
• SGAA (Simulation Gene Activation Agent)
• Vibrio Fischeri Simulation Program
• DB Agent (to access BioSPICE Data Warehouse
  connecting to SwissProt, KEGG, Enzyme,...)
• BioSPICE Monitor
• Contributed (not validated) SW also available (eg
  Charon, Pathway Builder, SBW Gateway,..)

37

• .

38
(No Transcript)
39
(No Transcript)
40
The ultimate reason for Systems Biology...

• Biologists observe things that cannot be
  explained. Theorists explain things that cannot
  be observed.
• attributed to Aharon Katchalsky
• by George Oster
• Thank you for your attention!

41
Thanks for your attention!

LMUs Old Physics Building