Transcript PowerPoint-Präsentation

SABIO-RK
Making Reaction Kinetics Data
Accessible
Martin Golebiewski
EML Research, Heidelberg, Germany
Biochemical Pathways
Enzyme Kinetics
Calculation of reaction rates:
e.g. Michaelis-Menten Kinetics
Kinetic Law:
V  S 
v m
K m  S 
Kinetic Parameters:
Vm = Maximum velocity
Km = Michaelis Constant
Kinetic Data: Database Resources
BRENDA
http://www.brenda.uni-koeln.de/
Comprehensive list of kinetic parameters based on literature information
UniProt Knowledgebase/Swiss-Prot
http://www.expasy.org/
Represents kinetic parameters in comment lines of proteins (new feature)
BioModels Database
http://www.ebi.ac.uk/biomodels/
Stores published mathematical models of biological interests
Why Another Kinetics Database ?
• Kinetic parameter values depend on environmental conditions
(temperature, pH, concentrations, experimental assay, etc.)
• Parameters only interpretable with their corresponding kinetic laws
• None of the databases links experimental kinetic parameter data for
single reactions to complete sets of information comprising
- Kinetic Law for the reaction rate
- Environmental conditions
- Concentrations of reactants and modifiers
- Data source (original publication)
- Organism, tissue and cellular location
• Kinetic Data must be easily accessible and interchangeable
We need an accessible database that merges this information
for published kinetic data
SABIO-RK: Database population
SABIO-RK describes Reaction Kinetics and is an extension of
SABIO (System for the Analysis of Biochemical Pathways)
SABIO
KEGG
UniProt
Extraction
Other DBs
Enzymes
Organisms
Reaction
Reactants
Pu
b
Pathways
SABIO-RK
Concentrations
Reactants
Kinetic Law
Environment
Parameters
Kineti
c
Data
(publ.)
SABIO-RK: Database population
Workflow
Kinetic data manually extracted from literature by students
Input
interface
Information to
unify the data
Intermediate SQL database
Curators verify and
complete the data and
check it for consistency
Data structuring
SABIO-RK / SABIO
Integration into a relational database (Oracle)
SABIO-RK: Database population
Curation Problems
• Information missing in publications about assay proceedings,
experimental conditions or complete reactions (with all reactants)
• Synonymic or aberrant notations of compounds and enzymes:
- different nomenclature systems used
- commonly used trivial names vs. systematic names
- functional classification of enzymes vs. physical enzyme proteins
(e.g. only 1 EC-class for hundreds of protein kinases)
• Multiplicity of parameter units:
- mass or molar concentrations (mg/ml vs. mmol/l)
- different systems (e.g. katal, unit or mol/min for enzymatic activity)
- enzyme activities can be specific (unit/mg) or absolute (unit)
- different scales (e.g. M, mM, µM, nM)
SABIO-RK: Database population
Statistics
• 20 students extracting kinetic data from literature
• 3 curators verifying and inserting data into SABIO-RK
• 735 publications entered by students (7180 entries for 1900 reactions)
• Integrated in SABIO-RK:
~ 4500
Entries (38% with kinetic law equation)
~ 1000
Different reactions (in 170 organisms)
~ 300
Different enzymes (EC classes)
~ 11300
Kinetic parameters:
4014
3666
852
Km and similar
kcat and Vmax
Ki (inhibitor constants)
SABIO-RK
SABIO-RK: User interface
• Web-accessible and freely available for academic users
• Users can query SABIO-RK by specifying
- reactants, enzymes and/or pathways
- organisms, tissues and cellular locations
- experimental conditions (pH, temperature)
• Links to other databases:
- SwissProt (for protein information about enzymes)
- PubMed (to refer to original publication)
• System supports export of the selected kinetic data in SBML
(Systems Biology Mark-up Language) format
SABIO-RK: SBML output
Aldolase reaction
Glycolysis reactions
SABIO-RK: Future Perspectives
• Extension of search functions in the user interface
e.g. find pathway from one compound to another
• Expansion of the data export functions in the user interface
Information very limited that can be exported in SBML format
• Description of the detailed mechanism of the reactions
e.g. elementary steps with corresponding rate constants
• Annotation to other hierarchy levels (building ontologies)
e.g. integration of classification tools for chemical compounds
• Annotation to other databases and usage of controlled
vocabulary
e.g. BioModels, ChEBI, KEGG, UniProt/SwissProt
SABIO-RK: Future Perspectives
• (Semi-)Automatic extraction of kinetic data by text mining tools
Problems:
- Publications containing kinetic data have to be identified
- Most of kinetic data found in tables, graphs or figures
- Reaction formulas and kinetic law equations are mainly pictures
- Automatic annotation of compounds, parameters, etc. is difficult
Manual data extraction might help to find rules for text mining.
• SABIO-RK as platform for experimental kinetic data
- Scientists producing the data can directly enter it into SABIO-RK
- All the needed information can be given by the experimenters
- No data lost or questionable because of missing information
Summary
• SABIO-RK is a web-accessible database containing biochemical
reaction kinetics data for systems biologists and experimenters
• Merges general reaction information retrieved from other
databases and kinetic data manually extracted from literature
• High degree of interrelation (all necessary information is linked)
• Type of kinetics, modes of inhibition or activation and
corresponding equations are shown with their parameters,
measured values and experimental conditions
• Kinetic data can be exported in SBML file format
• http://sabio.villa-bosch.de/SABIORK
Acknowledgements
Scientific Databases and Visualization Group:
• Renate Kania and Ulrike Wittig (database curators)
• Andreas Weidemann (programming: input interface, SQL database)
• Olga Krebs (database administration)
• Saqib Mir (programming: user interface, web-services)
• Isabel Rojas (programming and trying to organize all this)
• all our students entering the data
For helpful discussions:
• Ursula Kummer, Sven Sahle, Ralph Gauges (BCB-Group)
• Rebecca Wade and Matthias Stein (MCM-Group)
Thank you !