Tutorial:Network Loading And ID Mapping

From OpenTutorials
Jump to: navigation, search

Slideshow Network Loading and ID Mapping (20 min)
Handout Network_Loading_And_ID_Mapping.pdf (7 pages)

Tutorial Curators Mike Smoot

Cytoscape is an open source software platform for integrating, visualizing, and analyzing measurement data in the context of networks. This tutorial will introduce you to:

  • Searching internet interaction databases with query terms.
  • Mapping Identifiers of different types to networks.
  • Finding your query terms in the downloaded network.

Setup: Load necessary plugins

  • Start Cytoscape.
  • Under the Plugins menu, select Manage Plugins.
  • In the search dialog type "PSICQUIC".
  • Find the "PSICQUICUniversalClient" plugin and if it is not already installed, click the Install button.
  • Wait for the plugin to install.
  • Repeat for the plugin "EnhancedSearch".
  • Repeat for the plugin "CyThesaurus".
  • Close the Plugin Manager.

Search for a Network

  • Under the File menu, select Import → Network from Web Services....
  • In the dialog that pops up choose PSICQUIC Universal Web Service Client from the Data Source combo box. (Other data sources may work for this tutorial as well, so feel free to try one of those instead.)
  • In the Query field, copy the following gene IDs: YKR026C, YGR218W, YGL097W and then click the Search button.

(This example is for yeast, but any species should work. Likewise, any ID type should work. However, different PSICQUIC service providers may deal with some IDs or species better than others, so your mileage may vary!)
Psicquic query dialog.png

  • After a few seconds you should see a new dialog pop-up after the initial search has happened. Click Yes to indicate that you'd like to create a new network.
  • The Search Result dialog will show you a list of service providers which check boxes next to those that have query results to provide. Uncheck all except the IntAct service.

Psicquic query result.png

  • Click OK.
  • Click OK in the Import Result box as well.

Psicquic import result.png
(The other option, Merge will pop up Cytoscape's Network Merge utility and will attempt to merge the specified networks.)

You should now see two networks created: an overview network and the network itself. For this tutorial, you can safely ignore the overview network. One problem with the network that was loaded is that the node identifiers are most likely not the search terms you entered. Instead what you see are identifiers provided by the PSICQUIC service provider based on how they interpreted your search terms. In addition to node identifiers, the PSICQUIC plugin also provided several attributes for the network, among which are alternative identifiers for the nodes. Our next step will be to use these alternative identifiers to map back to our original identifiers.

Map Identifiers

This section will demonstrate how to use the CyThesaurus plugin to create a new attribute with an identifier of a specific type based on an existing identifier of a different type.

  • Under the Plugins menu, select CyThesaurus
  • Before creating a mapping, we need to configure the data source for CyThesaurus. Click the ID Mapping Resources Configuration button at the bottom of the CyThesaurus dialog.
  • Select the Web Services entry in the "Sources of ID Mapping" tree on the left side of the dialog.
  • In the dialog that pops up, select the "BridgeDB web service" option, and in the next combo box that appears, choose the base URL appropriate for the species you're working with. Since our search terms were yeast genes, select http://webservice.bridgedb.org/Saccharomyces cerevisiae. Click OK.

(If the species you're working with doesn't appear in the list, try other web service options.)

  • Now ensure that the proper web service is checked in the "Sources of ID Mapping" tree on the left side of the dialog.

Cythesaurus mapping resource dialog.png

  • Click the Save current resources as default at the bottom of the dialog and then click the Close button.
  • You should now be back to the main CyThesaurus dialog.
  • In the top section of the dialog select the network you wish to work with. This should be named after the database that generated the network (e.g. "MINT" or "IntAct").
  • In the "Select source attribute IDType(s)" section, click on the "Key Attribute" column. You should see a choice box listing the available attributes for the network.
  • Select "PSI-MI-25.uniprotkb.top". This action will trigger CyThesaurus to guess at the type of the attribute and in the column to the left of "PSI-MI-25.uniprotkb.top" you should see "[Uniprot/TrEMBL]" get selected. You may select additional or alternative ID types by clicking on this field and checking the appropriate boxes. For this tutorial, "[Uniprot/TrEMBL]" is sufficient.
  • In the "Select destination attribute/IDType(s)" section, click the "Target ID Type" field. Choose "Ensembl yeast" from the list.
  • This will create a new attribute, however the type of attribute will be a List of strings, rather than the single string we'd like. To change this click on the "All target ID(s) or first only?" column in the "Select destination attribute/IDType(s)" section. Choose "Keep the first target ID only". This will create a simple string attribute with first synonym that CyThesaurus finds.

Cythesaurus mapping configured.png

  • Click OK.
  • You should now see a new attribute available in the attribute browser called something like "Ensembl Yeast".

Find the nodes you care about!

The final step in this tutorial is to find your query terms in the network that you just loaded. It is possible to simply scroll through the attribute browser to find your terms, however it is much more efficient to use the EnhancedSearch plugin.

  • Simply type one or more of your search terms into the "Enhanced Search:" text field in the tool bar at the top of the main Cytoscape window.
  • Enter your original terms YKR026C, YGR218W, YGL097W and you should see 3 nodes get selected in your network!

Network loaded ids mapped result.png

Possible Next Steps

Now that you have a network with the nodes you care about identified, there are many ways to proceed.

  • Import attributes using the "Ensembl Yeast" attribute as a key.
  • Analyze the topology of the network using the "Network Analyzer" plugin.
  • Grow the network by searching other databases for the same query terms and merging the results into your existing network.
  • Load attributes (e.g. expression measurements) and then use the VizMapper to create an informative visualization of the network.