Limits...
Combining computational models, semantic annotations and simulation experiments in a graph database.

Henkel R, Wolkenhauer O, Waltemath D - Database (Oxford) (2015)

Bottom Line: The introduced concept notably improves the access of computational models and associated simulations in a model repository.This has positive effects on tasks such as model search, retrieval, ranking, matching and filtering.Furthermore, our work for the first time enables CellML- and Systems Biology Markup Language-encoded models to be effectively maintained in one database.

View Article: PubMed Central - PubMed

Affiliation: University of Rostock, Department of Computer Science, Albert-Einstein-Straße 22, D-18059 Rostock, Germany, Department of Systems Biology and Bioinformatics, University of Rostock, Ulmenstrasse 69, 18057 Rostock, Germany and Stellenbosch Institute for Advanced Study (STIAS), Wallenberg Research Centre at Stellenbosch University, Stellenbosch 7600, South Africa ron.henkel@uni-rostock.de.

Show MeSH
Architecture of our graph database. Data from different models, simulation descriptions or ontologies are imported using format-dependent importers. Each import undergoes a post processing afterwards. The stored graph and index structures are available via two retrieval interfaces: Cypher and an adaption of Henkel et al. (10). Both are based on RestAPIs. The data itself are stored in a Neo4J graph database.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4352687&req=5

bau130-F5: Architecture of our graph database. Data from different models, simulation descriptions or ontologies are imported using format-dependent importers. Each import undergoes a post processing afterwards. The stored graph and index structures are available via two retrieval interfaces: Cypher and an adaption of Henkel et al. (10). Both are based on RestAPIs. The data itself are stored in a Neo4J graph database.

Mentions: We implemented the graph-based storage according to the architecture depicted in Figure 5. The Neo4J (http://www.neo4j.org/) database stores model files, simulation descriptions and model-related information in a graph manner. The retrieval engine is based on the ranked retrieval described in Henkel et al. (10). It allows users to access the data in the database, and retrieve ranked lists of results for their text queries. Queries are resolved using the Lucene framework (http://lucene.apache.org/core/), and ranked based on predefined similarity features. The data import pushes different data formats, including model code, simulation experiment descriptions and ontologies, into the graph database. Afterwards a post-process takes care of linking the added data of different domains.Figure 5.


Combining computational models, semantic annotations and simulation experiments in a graph database.

Henkel R, Wolkenhauer O, Waltemath D - Database (Oxford) (2015)

Architecture of our graph database. Data from different models, simulation descriptions or ontologies are imported using format-dependent importers. Each import undergoes a post processing afterwards. The stored graph and index structures are available via two retrieval interfaces: Cypher and an adaption of Henkel et al. (10). Both are based on RestAPIs. The data itself are stored in a Neo4J graph database.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4352687&req=5

bau130-F5: Architecture of our graph database. Data from different models, simulation descriptions or ontologies are imported using format-dependent importers. Each import undergoes a post processing afterwards. The stored graph and index structures are available via two retrieval interfaces: Cypher and an adaption of Henkel et al. (10). Both are based on RestAPIs. The data itself are stored in a Neo4J graph database.
Mentions: We implemented the graph-based storage according to the architecture depicted in Figure 5. The Neo4J (http://www.neo4j.org/) database stores model files, simulation descriptions and model-related information in a graph manner. The retrieval engine is based on the ranked retrieval described in Henkel et al. (10). It allows users to access the data in the database, and retrieve ranked lists of results for their text queries. Queries are resolved using the Lucene framework (http://lucene.apache.org/core/), and ranked based on predefined similarity features. The data import pushes different data formats, including model code, simulation experiment descriptions and ontologies, into the graph database. Afterwards a post-process takes care of linking the added data of different domains.Figure 5.

Bottom Line: The introduced concept notably improves the access of computational models and associated simulations in a model repository.This has positive effects on tasks such as model search, retrieval, ranking, matching and filtering.Furthermore, our work for the first time enables CellML- and Systems Biology Markup Language-encoded models to be effectively maintained in one database.

View Article: PubMed Central - PubMed

Affiliation: University of Rostock, Department of Computer Science, Albert-Einstein-Straße 22, D-18059 Rostock, Germany, Department of Systems Biology and Bioinformatics, University of Rostock, Ulmenstrasse 69, 18057 Rostock, Germany and Stellenbosch Institute for Advanced Study (STIAS), Wallenberg Research Centre at Stellenbosch University, Stellenbosch 7600, South Africa ron.henkel@uni-rostock.de.

Show MeSH