Identifying pattern in reaction networks of computational models
0 likes499 views
The document discusses methods for calculating similarity between biological simulation models, emphasizing annotation-based feature extraction and structure-based similarity techniques. It outlines what constitutes a simulation study in computational biology, including necessary data and model-related links. Additionally, tools like PARSEMIS for pattern detection are introduced to enhance model retrieval based on shared patterns.
Identifying pattern in reaction networks of computational models
- 1. http://sems.uni-rostock.de Identifying pattern in reaction networks of computational models Fabienne Lambusch, Rebekka Alm, Ron Henkel, Dagmar Waltemath October 2015 | COMBINE, Salt Lake City, US
- 2. How can we calculate the similarity between biological simulation models? 2
- 3. How can we calculate the similarity between biological simulation models? 3 thematic sets of models arbitrary models versions of models
- 4. 4 Data Jose-Juan Tapia @ COMBINE2015
- 5. 5 Data ● What constitutes a simulation study in computational biology? – Publication – Experimental data – Model file(s) – Simulation setup(s) – Simulation data – Visualisations – Model assumptions – Operating Procedures
- 6. 6 Data ● What constitutes a simulation study in computational biology? ● How are model-related data linked amongst each other?
- 7. 7 Data MASYMOS ● What constitutes a simulation study in computational biology? ● How are model-related data linked amongst each other? ● How can we represent these links on the storage layer? Henkel et al (2015)
- 8. 8 Model similarity Annotation-based ranked retrieval Model Publication Annotation Person Simulation Show me models by Tyson, dealing with the Cell Cycle and simulating concentration of cdc2! Ron Henkel Henkel (2010)
- 9. 9 Model similarity MASYMOS Ranking Henkel (2010) More descriptive model features from annotations Structure-based similarity scores
- 10. 10 Model similarity Annotation-based feature extraction Rebekka Alm GO (~11.000 references) | SBO (~13.000) | ChEBI (6.000) Adapted from: Courtot et al.: Controlled vocabularies and semantics in systems biology. In: Molecular Systems Biology 7 (2011), Nr. 1. Alm (2015) JBMS Number of reactant SBO: 0000167 SBML models from BioModels Database (R25) Rebekka Alm
- 11. 11 Model similarity Annotation-based feature extraction Adapted from: Alm (2015) JBMS 5 Features 15 Features GO GO:0022411 (depth 4) cellular component disassembly GO:0030163 (depth 5) protein catabolic process GO:0051726 (depth 5) regulation of cell cycle GO:0065009 (depth 3) regulation of molecular function GO:0071822 (depth 5) protein complex subunit organization GO (~11.000 references) | SBO (~13.000) | ChEBI (6.000) SBML models from BioModels Database (R25)
- 12. 12 Model similarity MASYMOS Henkel (2011) BMC SysBiol Alm (2015) JBMS Which models contain feedback loops? How many models contain this submodule? Which models have 'ATP' as product? Structure-based similarity scores CellCycle Models x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x BioModelsMorefeaturesRanking
- 13. 13 Structure-based similarity Fabienne Lambusch Models are similar, if they share (many) common subgraphs / patterns / motifs
- 14. 14 Structure-based similarity Pattern detection Parallel and Sequential Mining Suite (ParSeMiS) Find patterns which occur in at least X of the models ● Input: SBML models from MASYSMOS ● Output: List of patterns and frequencies ParSeMiS: https://github.com/timtadh/parsemis
- 15. 15 Structure-based similarity Pattern Frequency of occurrence (560 models, threshold 350 models)
- 17. 17 Structure-based similarity Clustering x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x CellCycle Models? x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x
- 18. 18 Summary Available pattern for search Set-dependent features Ranked model retrieval
- 19. Thank you for your attention. 19 Fabienne Lambusch | Martin Scharm | Dagmar Waltemath | Mariam Nassar Tom Gebhardt | Martin Peters* (all SEMS) | Vasundra Touré (SBGN-ED+) Ron Henkel (de.NBI-SYSBIO) @SemsProject 2013 2015 * did all these cool graphics http://sems.uni-rostock.de
- 20. Additional slide: Resources 20 @SemsProject MASYMOS Paper: http://database.oxfordjournals.org/content/2015/bau130 Software demo: http://graphdb-review.sems.uni-rostock.de:17474/browser/ Homepage: https://sems.uni-rostock.de/projects/masymos/ Annotation-based feature extraction Paper: http://www.jbiomedsem.com/content/6/1/20 Code: https://bitbucket.org/ronhenkel/masymos Homepage: https://sems.uni-rostock.de/projects/masymos/model_feature_extraction Structure-based similarity Paper: on the way… Homepage: https://sems.uni-rostock.de/projects/masymos/model_feature_extraction