Make Sure you are using Cytoscape 2.8.3!
The plugins used in this tutorial are listed below and can be installed in Cytoscape by Pplgins->Manage Plugins. Type the name of the plugin, search for it, select a version that works with Cytoscape 2.8.3 (there is a little green tick mark beside the plugin name indicating this) and click install.

  • GPML plugin
  • Cytargetlinker plugin
  • Cythesaurus plugin
  • MiMI plugin

Outline

Click on the figures on this page to see it in full size.


Tutorial 1: Protein-protein interactions in STRING

Step 1: Get data from STRING

When sorting the genes with a p-value < 0.05, A2m is the gene with the highest log fold change (5.2). That means that the gene is turned on in the animals fed with the high dose. In this exercise, we are going to further investigate the role and interaction partners of A2m.

  1. Go to the STRING database and search for A2m in Rat.
  2. The ten most common interaction partners are shown in the protein interactions network, see Fig 1a.
  3. When clicking on the “+” below the network, the next 10 neighbors are added, see Fig 1b
  4. Save the “Text Summary” of this network containing 21 nodes, see Fig 1c.

 

Fig 1a PPi network for Alpha-2-M Precursor protein with first ten neighbours Fig 1b PPi network for Alpha-2-M Precursor protein with additional neighbours Fig 1c Click on save and scroll down to save the network as a tab delimited text file
Fig 1a. String network for Alpha-2-macroglobulin Precursor (Alpha-2-M) protein with first ten neighbours Fig 1b. String network for Alpha-2-macroglobulin Precursor (Alpha-2-M) protein with additional neighbours Fig 1c. Click on save and scroll down to save the network as a tab delimited text file

 

Step 2: Open network in Cytoscape

Start Cytoscape 2.8.3. The downloaded file from STRING can be opened as a network in Cytoscape:

  1. File -> Import -> Network from Table, see Figure 2a
  2. Click on “Show Text File Import Options” and select “Transfer first line as attribute names” to make sure the header line is not used as a node.
  3. Specify the first column as a source interaction and the second as target interactions.
  4. Click on Import, see Figure 2b
  5. Cytoscape offers different layout algorithms. Try out which layout seems most intuitive for you (Fig 2c). Click on Layout -> Cytoscape Layout, JGraph Layouts or yFiles, see Fig 2d (shows how the network would look when Force Directed Layout is applied).

 

Fig 2a Import Network from Table Fig 2b File import dialog Fig 2c Try different layouts Fig 2d Imported network with force-directed layout as an example
Fig 2a Import Network from Table Fig 2b File import dialog Fig 2c : Try different layouts Fig 2d Imported network with force-directed layout as an example.

Step 3: Analyse and visualize network properties. Now we are going to look at network properties that were discussed in the presentation this morning. In Cytoscape we can use the NetworkAnalyzer plugin which calculates all the properties in the network.

  • Click on Plugins -> Network Analysis -> Analyze network -> Treat network as undirected.<(Fig 3a & 3b)
  • Check the distribution for node degree and shortest path lengths, see Fig 3c & 3d.
  • Click on “Visualize Parameters” in the bottom of the dialog. Visualize the node degree and betweenness on the nodes in the network (see Fig 3e):
    • “Map node size to” -> Select “Degree”
    • “Map node color to” -> Select “Betweenness Centrality”
  • Now the node size tells you how many nodes are connected with the node and the color indicates how crucial the node is for the connectivity of the network, see Fig 3f.
Fig 3a Analyze network Fig 3b Treat network as undirected Fig 3c Node Degree Distribution
Fig 3a Analyze network Fig 3b Treat network as undirected Fig 3c Degree Distribution
Fig 3d Shortest path length Distribution Fig 3e Visualize Parameters Fig 3f Data Visualized Network
Fig 3d Shortest path length Distribution Fig 3e Visualize Parameters Fig 3f Data Visualized Network

Tutorial 2: Extending a biological pathway with regulatory interactions

Step 1: Open a WikiPathways pathway in Cytoscape
You can either download the gpml file from WikiPathways, copy/paste a pathway from PathVisio into Cytoscape or use the webservice to download the pathways from within Cytoscape. In this practical we are going to use the webservice.

  1. Make sure that the GPML (v 1.4) and the CyTargetLinker (v 2.1) plugin are installed. If not, try to download them, if it takes too long depending on internet connection ask one of the instructors for a USB stick.
  2. Go to File -> Import -> Network from Webservice, see Fig 4a
  3. Select WikiPathways as the data source
  4. Search for the human “Statin” pathway, see Fig 4b
  5. Click on the “Statin Pathway”
  6. The pathway should show up as a network in Cytoscape, see Fig 4c

 

Fig 4a: Load a network from a webservice Fig 4b: Download the human Statin pathway from WikiPathways Fig 4c: Pathway is shown as network in Cytoscape
tutorial2_1 tutorial2_2 tutorial_2_7

 

Step 2: Extend the pathway with regulatory interactions
Now we want to use CyTargetLinker to integrate regulatory interactions including microRNA-gene, transcription factor target and drug-target interactions. Those interactions are provided in so called RINs (Regulatory Interaction Networks). Please download this zip file and extract the content.

  1. Go to Plugins -> CyTargetLinker -> Load Regulatory Interaction Networks, see Fig 5a
  2. The node attribute “GeneID” contains the biological identifier which is needed by CyTargetLinker
  3. In the “Select RINs” section, click the browse button and select the directory that you downloaded before (called hsa-rins)
  4. Change the integration direction to “Add regulators” so only microRNAs, transcription factors and drugs are added.
  5. Click on “Ok”
  6. We will use the all the provided RINs which include validated microRNA-target interactions from miRecords and miRTarBase, TF-gene interactions from ENCODE data (split up in proximal and distal regulation) and drug-target interactions from DrugBank
  7. The resulting network contains the original nodes from the network in white and the new regulators as pink rounded rectangles, see Fig 5b. Some of the nodes might be disconnected because the pathway is not fully connected or contains additional annotation elements.

 

Fig 5a: CyTargetLinker dialog settings Fig 5b: Extended network in Cytoscape, the edge colors indicate from which RIN the interaction originates.
tutorial2_4 tutorial2_5

 

Step 3: Change the visual style to show the biological type
Since we integrated several different regulators, it is useful to change the color based on the biological type. The RINs provide a node attribute “biologicalType” that categorizes the regulators.

  1. In the control panel on the left side, click the “VizMapper” tab and change the mapping for the “node color”, see Fig 6a and b
  2. Drugs will be visualizes in green, TFs in purple and microRNAs in red.
  3. This network can be considered the starting point to visualize experimental data and find new hypothesis. CyTargetLinker is not restricted to these resources and can use other regulatory interactions as well.

 

Fig 6a: VizMapper node color settings Fig 6b: Extended network in Cytoscape with different colored regulators.
tutorial2_6a tutorial2_6b

 


Tutorial 3: Find interaction data and visualize your data on the network

Step 1: Using the MiMI plugin to find interactions
The file top-genes.txt (right click – save as) contains the top up and down regulated genes (pValue < 0.05) in the dataset. We will use this gene list to find molecular interactions between the genes and their first neighbors.In Cytoscape we can use the MiMI (Michigan Molecular Interactions) plugin to search for interaction data from several different databases.

  • Go to Plugins -> MiMI plugin -> Query, see Fig 4a
  • Select tab “From File”
  • Browse “Load Gene File” to select the top-genes.txt file, see Fig 4b
  • Select Rattus Norvegicus as a species and leave the other settings as default, see Fig 4c.
  • The resulting file shows the query genes as grey diamonds and the first neighbors as pink circles, see Fig 4d.

 

Fig 4a Start MiMI plugin Fig 4b Load top-genes.txt Fig 4d The extended gene network
mimi1
mimi2
Fig 4c Set Query Parameters
mimi3
mimi4.PNG

 
Step 2: Map Identifiers, Import Data and Visualize
The next step is to import gene expression data and map it to the nodes of our network and visualize the data. We will use the same dataset here as we used for Pathway Analysis yesterday. The first column of the dataset contains Ensembl identifiers for the genes measured in the dataset. But, the network we generate in the previous step using the MiMI plugin, creates a network in which the nodes are identified by Entrez Gene identifiers.
So, in order to map the data onto the network we need to first map the identifiers present in the dataset and the network to each other. This we do using the Cythesaurus plugin.

Map identifiers:

  • Open Cythesarus plugin, see Fig 5a.
  • Set Identifier Mapping Resource, see Fig 5b and 5c.
  • Set source and target identifier systems, see Fig 5d.

You should get a pop-up confirming that the identifiers have been mapped, see Fig 5e.
 

Fig 5a Open Cythesarus plugin Fig 5b Set Identifier Mapping Resource Fig 5c Select Identifier Mapping Resource Fig 5d Set source and target identifier systems
cythesaurus-1 cythesaurus0 cythesaurus1 cythesaurus2
Fig 5e Confirmation pop-up
cythesaurus3

 
Import Data

  • File -> Import -> Attributes from table.., see Fig 6a
  • Set Parameters

You should see a pop-up box confirming the import of data.
 

Fig 6a Import Attributes Fig 6b Set Parameters Fig 6c Confirmation pop-up
data1 data2 data3

 
Visualize Data

  • Select VizMapper in side panel and set parameters, see Fig 7a
  • Double click on the gradient in the side panel to configure it, see video tutorial

The resulting colored network is shown in Fig 7b.
 

Fig 7a Setting VizMapper parameters Fig 7b Data Visualized Network
vizmapper final-network

 
Video Tutorial : How to set the gradient visualization