Published on September 24, 2007
BioPathways SIG Report : BioPathways SIG Report 5th BioPathways Consortium Meeting at ISMB’03 The BioPathways Consortium: The BioPathways Consortium Scope : Computational Biology and Informatics related to Metabolic and Signaling Pathways, Regulatory Networks, and Protein Interactions Mission : Foster development of computational pathways biology Goals : Scientific community buildup, standards recommendation, public outreach, industry-academia collaboration support, fund raising, coordination with other groups andamp; consortia Means : Open forum of interested participants (academics, pharmas, biotechs, software vendors) BioPathways Consortium Activities: BioPathways Consortium Activities Scientific Meetings : 1st Meeting at ISMB 2000 -andgt; Work Groups 2nd Meeting at PSB 2001 -andgt; First results on evaluation of pathways representations 3rd Meeting – Satellite Meeting of ISMB 2001, Copenhagen -andgt; Focus on ontologies and pathways reconstruction 4th Meeting – BPCM’02@Edmonton -andgt; Focus on reconstruction and systems biology 5th Meeting – BPCM’03 Workgroups : formalisms, reconstruction, text-mining, visualisation ISCB SIG Several sponsors (pharmas, biotechs, IT companies) Web site - http://www.biopathways.org Computational Biology of Pathways Evolution of the Field: Computational Biology of Pathways Evolution of the Field Coming of age Shift from genomics to ...*.omics Stronger coupling of theoretical and experimental approaches Explosion of papers andamp; more selective refereeing in biological journals Back to more traditional roles for industry and academia Major themes : Modeling Reconstruction Simulation Analysis : structural and dynamic properties, design principles, definition and identification and evolution of modules... Database and tools development Back to basics with a new perspective, eg metabolism This year’s meeting: This year’s meeting 21 speakers Sessions : Regulation and interactions on a systems scale Function and evolution of metabolic networks Ontologies, databases and data integration Contributed session : pathway tools Open Discussions 90 official participants (+ BOSC sporadic participants) Regulation and interactions: Regulation and interactions Modularity (Segal, Przulj) Different levels of molecular organization : Regulatory modules (graphical probabilistic models) Protein interaction modules (graph theoretical models) Module conservation across species Cleaner functional relationships General design principles Modules still require careful definition and integration across different levels Biological relevance Statistical robustness Stochasticity Stochastic modeling (Tian) Stochastic inference methods (Segal) Regulation and interactions: Regulation and interactions Prediction of individual properties (Zhang, Przulj, de Lichtenberg, Segal) Functional annotation Physical interactions Lethality Synthetic lethality …using sets of (other) properties Validation Cross validation ROC curves Biological validation Open Discussion : Regulation and Interactions: Models What kinds of models to learn? Relationships between different levels of network description : structural features, logical view, detailed dynamics Are proteins the right level of abstraction ? What about domains ? Data How to combine different data types? How to move to a multi-species scale? Functional categories are consistently used as reference datasets to validate predictions…can we start upgrading our view of function ? Robustness of methods vs incompleteness/error in experimental data Why is there so little overlap between interaction datasets ? Open Discussion : Regulation and Interactions Open Discussion : Regulation and Interactions: Results What kind of predictions can methods suggest? What kinds of experiments? Design principles in different networks: ppi, regulation etc (e.g. modules) – and different data (expression, cis-regulatory motifs). Is there a link between them? Can we/should we organize a competition for protein interaction prediction ? Open Discussion : Regulation and Interactions Metabolism: Metabolism Comparative approach (Shah, dos Santos) To fill in the gaps in the reconstruction of a static metabolic network To explain major phenotypic differences between genetically close species using FBA (steady state dynamics) Multi-cellular and multi-organism models (dos Santos) Dynamical models (Almeida, Tomita, dos Santos) Full dynamical model : difficult to reconstruct and study Steady-state approximations Intermediate models : eg S-Systems : easy connection to graph topology, decoupling of enzyme properties inference Coupling of small differential equation model with steady-state Whole organism metabolism : the E. coli example(Tomita, Keio University): Whole organism metabolism : the E. coli example (Tomita, Keio University) Analysis of cellular metabolites: CE/MS measurements for charged molecules, LC/MS for neutral ones Inference of metabolic networks from time series of metabolite concentrations In vivo dynamic analysis / Pulse chase with labeled isotopes Expression profiling and protein 2D-gels Systematic in vitro kinetics of individual enzymes (full kinetic data by end of year) Single deletion mutant library (for each single gene) : metabolome for each mutant Protein-protein interactions : His-tagged/MS or RNA-tagged + PCR Currently : 10000 interactions with ~2600 proteins Protein localization analyis GFP-tagged libary for all 4000 E. coli genes From first analysis most proteins seem to be uniformly distributed in the cell Open discussion : Metabolism : Open discussion : Metabolism What are the goals in our study of metabolism ? Design principles ? « Modules » ? Forward-engineering of bacteria ? In multicellular organisms ? Function(s) of metabolism ? How much is actually known about metabolism ? Difference between « potential » and actually occurring pathways ? Gap between canonical pathways and condition-specific pathways What is the power of comparative approaches ? Variability of metabolism in prokaryotes ? Interplay between regulation and metabolism ? How can the models be « integrated » ? Open discussion : Metabolism : Open discussion : Metabolism Experimental data : Timeline for availability What do we really need ? Data quality / reliability issues Importance of individual cell measurements ? Should we work with chemostats for basic research ? Drift of bacterial populations during prolonged metabolic experiments. Which models/formalisms for metabolism ? Static vs dynamic models ? Combination ? Validity of the well-stirred reactor approximation ? Space should be forgotten for now… Steady-state assumption ? Objective functions and optimality assumptions Database & ontologies: Database andamp; ontologies Exchange languages for dynamical models CellML (Haalsted/Nielsen) SBML (Hucka) Exchange formats for databases BioPAX (Luciano) OMG-LSR announcement (Markel) Pathways databases Patika (Dogrusoz ) Tools session: Tools session CADLIVE for constructing a large-scale biochemical network (Hiroyuki Kurata) GenMAPP and MAPPFinder 2.0: Tools for the Organization, Display, and Exchange of Pathway Information (Kam Dahlquist) Cell Illustrator (Eleanor Zhou) BioPACS: BioPathway Automatic Convert System for Genomic Object Net (Hiroshi Matsuno) A recursively decomposable pathway DB: annotating processes and sub-processes (Ken Ichiro Fukuda) Acknowledgements: Acknowledgements Joanne Luciano Eric Neumann Speakers ! ISCB, Stephanie Hagstrom, Michael Gribskov Air-France, All Nippon Airways, Qantas Qantas baggage service Chris Dagdigian for flawless wireless Steven Leard See you next year in Glasgow !