Extend a set of microRNAs with regulatory information
Created with Cytoscape 3.6.1 and CyTargetlinker 4.1.0
CyTargetLinker provides a quick and extensive enrichment of networks in Cytoscape. The visualization options allow biological interpretation of complex regulatory networks in a graphical way.
In this tutorial a step-by-step explanation is given on how to extend a list of microRNAs with predicted and validated microRNA-target interactions (MTIs) in Cytoscape versions 3.6+ using the new CyTargetLinker 4.0+ app
- Step 1: Import a set of microRNAs into Cytoscape 3.6+
- Step 2: Download MTI link sets
- Step 3: Built regulatory network in Cytoscape
- Step 4: Adapt visualization options
Step 1: Import a set of microRNAs into Cytoscape 3.6+
First you will need to create a biological network in Cytoscape 3.6+ containing your microRNAs of interest. The datafile should contain at least two columns, (i) the microRNA name and (ii) the corresponding miRBase accession number. An example datafile, in which two microRNAs known to be involved in obesity are listed, can be found here. Open the datafile containing the microRNAs in Cytoscape via File Import Network File. The default settings, should already recognize the source node and a node attribute column (see Figure 1). After clicking ‘OK’, you will receive the following warning: “No edges will be created in the network; the source column is not selected. Do you want to continue?”. Click ‘Yes’.
The resulting biological network of two microRNAs from the example data, as shown in Figure 2.
Step 2: Download MTI link sets
A link set is a network containing additional information and links about nodes in the network. This could be regulatory interactions like microRNA-gene interactions. The networks are stored in XGMML (the eXtensible Graph Markup and Modeling Language) format, which is supported by Cytoscape. Each link set contains interactions consisting of two nodes, source and target, connected through one directed edge. Several link set for different use cases, species and interaction types are provided on our website. CyTargetLinker is not limited to the provided link sets. A user can also easily create their own link sets.
Before starting CyTargetLinker you need to download the link sets of interest and store them in a directory. In this example, we will use microRNA-gene interaction link sets from miRTarBase (validated interactions) and TargetScan (predicted interactions). Download the latest human link sets for both databases from here. In this tutorial, I used miRTarBase version 7.0 and TargetScan version 7.2. After downloading the two link set, unzip the files in a new directory.
Step 3: Built regulatory network in Cytoscape
Install the CyTargetLinker app via the app manager.
Open the CyTargetLinker in Cytoscape via Apps CyTargetLinker plugin Extend network. Thereafter the CyTargetLinker selection panel is displayed (Figure 4).
- Select the user network. This is the biological network you want to extend with regulatory information. Here we select the human microRNAs (‘Sheet 1’ if you haven’t renamed the network yet).
- Select the network attribute. This is the column containing the identifier which is supported by CyTargetLinker. The supported identifiers are listed in overview of the link sets. For the microRNAs we recommend to use the miRBase accession numbers.
- Select the directory containing the species specific link sets which you downloaded in step 2.
- Select the direction of the interaction. The user can decide how he wants to extend the network, i.e., 1) add regulators (interactions from target to source), 2) add targets ( interactions from source to target) or 3) both (default). Here we add only the targets of the microRNAs.
Click ‘OK’ en thereafter select the link sets containing microRNA-target information from miRTarBase and TargetScan (Figure 5).
The extended network is now visualized in a new Cytoscape network, as shown in Figure 6. Go to Apps -> CyTargetLinker -> Show result panel. You can now select the extended network and see how many interactions were added from each link set (942 from miRTarBase and 2,997 from TargetScan). The edge color indicates from which link set the interaction was added (red for miRTarBase and blue for TargetScan).
Since the default layout (forced-directed) is not ideal for this specific extended network, you can try other layouts (e.g. the yFiles organic layout, see Figure 7). This gives you a better idea about how many shared targets the two microRNAs have.
Step 4: Adapt visualization options
At any point can you hide certain link sets by using the ‘Show/Hide’ dropdown box in the results panel. You can also change the edge color used for a specific link set in this panel.
A very useful feature for large networks with multiple link sets containing the same interaction type, you can use the ‘Overlap threshold’ functionality. If you select ‘2’, then you will only see those interactions that are present in at least 2 link sets.
If you try this in the tutorial network, you will only see 454 microRNA-gene relationships (2 edges between each of them – from each database one, see Figure 8). Those interactions are the only ones that are present in both databases.