Network building with Cytoscape App queries to the
                                  BioGateway 3.0 triple store

                                                      Stian Holmås
                                                    Vladimir Mironov
                                                      Martin Kuiper
                                 Norwegian University of Science and Technology (NTNU)
                                 {stian.holmas; vladimir.n.mironov; mtrkuiper} @gmail.com

                   Abstract. We have developed the BioGateway App, a Cytoscape plugin that sup-
                ports the building of a wish list of questions that are translated to SPARQL queries,
                which are launched against the BioGateway server. We demonstrate the functionality
                of the BioGateway App both in simple and more complex use cases, taken from our
                website https://www.biogateway.eu/examples/

                     Keywords: triple store, cytoscape app, network building.


                1.       Introduction

                1.1.     The BioGateway triple store

                The BioGateway (BGW) triple store [1, 2] was one of the first major RDF resources
                available with biological information. The use of SPARQL to query triple stores proved
                to be one of the major limiting factors for the wider acceptance of the Semantic Web
                technologies by systems biologists. Therefore, we now reach out to the relatively large
                user community of the biological network analysis platform Cytoscape through the de-
                velopment of the BioGateway App plugin.

                1.2.     The BioGateway Cytoscape App

                The main feature of the BioGateway App is the Query Builder, which supports the
                design of queries that are built from definitions of proteins or genes and a relationship
                to either an ontology term or another protein or gene. By adding additional query parts
                line by line, increasingly complex and restrictive or inclusive queries can be composed.
                The Run Query command converts these to native SPARQL queries that are launched
                against the BioGateway 3.0 SPARQL endpoint [3].

                1.3.     The biological data in the store

                The main information that is subject to the query is obtained from IntAct [4], Uni-
                ProtKB [5] and the Gene Ontology database [6]. To allow a user a special focus on
                gene regulation we also included several resources with regulatory relations of tran-
                scription factors and their target genes.




Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
2


2.        The domo: querying the store

2.1.      Autocomplete support for query building

The selection of terms of interest such as genes, proteins, ontology terms and relation
types is facilitated by an autocomplete function that is driven by a REST API of a
NoSQL database loaded with all the entity names and metadata from the BioGateway
server, to allow quick response times. To ensure compatibility between the App and the
BioGateway data, upon startup the App fetches an XML-based configuration file from
our webserver. This file contains the relation types and their URIs, default settings and
the default layout style for BioGateway graphs, as well as available metadata types and
the query constraints that this metadata enables, allowing some updates to the App
without requiring the user to reinstall it. User preferences set in the BioGateway tab of
the Cytoscape Control Panel are stored between sessions, such as the default query
constraints related to species, data sources, and additional selection criteria for querying
the BioGateway content.

2.2.      Query constraints

Next to specifying the results through the definition of restrictive queries, the Query
Constraints section of the BioGateway tab in the Control Panel allows additional con-
straints for specific relations, such as setting a minimum confidence score for Protein-
Protein Interactions, as provided by IntAct (Orchard et al. 2013). This control panel
also allows the selection of extra types of metadata to be loaded together with the re-
sults, but as this may significantly increase query time, this metadata can also be added
after the network is complete (Reload Metadata), so that it can be used for filtering and
display options.

3.        Scope of the demo

We will demonstrate the functionality of the BioGateway App both in simple and more
complex use cases, taken from https://www.biogateway.eu/examples/. While BGW
3.0.0 contains only human data, BGW 3.0.1 (release date expected in December 2019)
will include data for 25 best-studied eukaryotes.


4.        References
    1. Antezana, E. et al. (2009) BioGateway: a semantic systems biology tool for the life sciences.
       BMC Bioinformatics, 10, S11.
    2. BioGateway home page, https://www.biogateway.eu/#database
    3. BioGateway SPARQL endpoint: www.biogateway.eu/sparql-endpoint/
    4. Orchard, S. et al (2013) The MIntAct project—IntAct as a common curation platform for 11
       molecular interaction databases. Nucleic Acids Res., 42, D358—D363.
    5. Gene Ontology Consortium (2018) The Gene Ontology resource: 20 years and still GOing
       strong. Nucleic Acids Res., 47, D330—D338.
    6. Uniprot ref Author, F.: Contribution title. In: 9th International Proceedings on Proceedings,
       pp. 1–2. Publisher, Location (2010).