How to query HSC

  • Default query:
    By default all interactions in all categories are searched. Queries are not case-sensitive and search the complete database content that includes the query term (e.g. repop is sufficient to find everything concerning repopulation activity). This type of query is highly unspecific as it includes also information such as gene/protein name synonyms and author names from referenced literature. A search for e.g. Cd150 in 'all fields' lists not only all HSC-subtypes with Cd150 in its immunophenotype but also all results about the gene Cd150 being the synonym of genes Slamf1.

  • Specific query:
    The dropdown menu under 'field' allows searching within a specific type of data, e.g. only within genes/proteins, biological processes, immunophenotypes (to identify the HSC sub-type) etc.

  • Combined queries can be performed by using Refine query. By clicking on Refine query a second search field is opened. A drop-down menu allows combining the two queries with the Boolean operators AND, OR and NOT.

  • In case of search terms containing blanks, it is recommended to set the term into double quotes, e.g. "repopulation >16 weeks".

  • The character for wild-card is '%'. It can also be used in a double-quoted search term at every position. A search for % in 'all fields' finds all interactions shown in the database.

The search results are listed in the form of a table but can also be graphically visualized with a graphic tool that dynamically generates a network (see Graphical visualization of search results).

Visualization of data

On the graphical output all categories, such as immunophenotype, biological processes or genes/proteins are displayed as colour-coded square nodes which are combined by interactions (edges). The output is limited to 2000 interactions.

  • Pop-up windows show additional information to each interaction
    - Moving the cursor over edges opens a pop-up window with detailed information about the underlying interactions.
    - Clicking on an edge results in opening a new window containing the complete entry information.
    - The thickness of an edge correlates with the number of interactions found for the same subject, object and interaction type.

  • Access to information from external resources
    - Moving the cursor over nodes (boxes) opens a pop-up window with information about protein names, chemical compounds and definitions from Gene Ontology and BRENDA Tissue Ontology.
    - A single click on a box while holding the CTRL button opens a link to the external resources such as EntrezGene, Gene Ontology and KEGG.

  • Moving and clustering nodes
    For moving the nodes one has to click on the icon 'Move node' on the panel. The color of the respective node turns darker and the node can be moved by using the left mouse button. Adjacent nodes can be clustered by dragging the mouse pointer over the nodes of interest. The nodes can than be moved as a cluster.

  • Extending the graph for items of interest
    Double clicking on a node launches a new search extending the current information for all interactions that are associated with the respective node.

  • Search for nodes inside a graph
    In complex graphs nodes of interest might be difficult to see. Using the search field on the upper middle of the graph highlights terms of interest and places them in the middle of the graph page.

  • Zooming and moving of the graph
    For zooming the graph either the mouse wheel or the zoom function on the upper left of the graph can be used.
    For moving the graph the left mouse button has to be hold down and the mouse to be dragged in the appropriate direction.

  • Hide/show transitive interactions
    Transitive interactions between nodes which are not displayed in the search result can be visualized by clicking on the button "Show transitive interactions". i.e. if a search finds a-b and a-c, a interaction between b and c will also be shown, even if it is not found by your original search.

  • Graph download
    Diagrams can be downloaded as SBML, graphML or jpg files by clicking on the respective buttons on the lower left of the window.
    SBML: SBML files can be processed by numerous software tools (see SBML Software Guide). Instructive layout styles for Cytoscape users are e.g. yFiles organic, yFiles orthogonal, Cytoscape Force-Directed Layout, Cytoscape Spring Embedded and Cytoscape Edge-Weighted Force-Directed.
    GraphML: Users that want to edit the graph should use the graphML format. The file can be opened and edited with the freely available yED software (

Origin of the data

Data in HSC-Explorer are retrieved from scientific articles and are linked to the respective PubMed entries. All data are manually annotated by experienced biocurators.

Curation of data

  • General information
    includes the literature reference (PubMed-ID) and information about the organism (strain, gender and age) that was used for the experiments.

  • Experimentally verified biological information extracted from the publications is shown in the comment. This field includes in addition a short description of the methods used. Background information necessary for understanding but not experimentally verified in the "source" publication is marked as "cited information".

  • The structured information translates the biological findings into very basic information, that is the interaction between two elements. This is also known as subject-predicate-object structure. "Subjects" and "Objects" are linked to ontologies such as EntrezGene, KEGG, miRBase, Gene Ontology or CORUM, if possible. The structured information can be extended by a further text field containing information about important experimental conditions. In addition we indicate whether the result is obtained from in vitro or in vivo experiments.


Dr. Andreas Ruepp
Institute of Bioinformatics and Systems Biology/MIPS
Phone:  ++49-89-3187-3189
Fax:      ++49-89-3187-3585