Max-Planck-Institut für Informatik
max planck institut
informatik
mpii logo Minerva of the Max Planck Society
 

NetworkPrioritizer - Tutorial

Tutorial

This section presents a short tutorial about NetworkPrioritizer that covers the case study described in the Applications Note. To complete the tutorial, you should download the sample data archive (~ 120kb). Please note that this tutorial assumes that you are familiar with Cytoscape 2.8.3 to some extent.
Hint: You can click on the screenshots to see them in full size.

Once the archive is extracted and Cytoscape is running, you can import the network contained in the file CDnet.xgmml. Your Cytoscape desktop should then look similar to Fig. 1. The next step is to specify the nodes that should be used as seeds in the network analysis. You can do that by choosing NetworkPrioritizer -> Seed nodes -> Tag from file from the Plugins menu, as shown in Fig. 2. Select the file cd_associated.tpl that was contained in the sample data archive. NetworkPrioritizer should tell you then that "106 nodes have successfully been tagged as seed". Cytoscape's VizMapper can be used to indicate the seed nodes in the network view, as shown in Fig. 3.

Fig. 1: Cytoscape Desktop after importing CDnet.xgmml. Fig. 2: How to find the menu entry Tag from file. Fig. 3: Seed nodes are marked in red via Cytoscape's VizMapper.

The loaded network contains edges that represent protein-protein interactions and edges that represent functional similarities. An edge attribute weight already exists, but it is set to 1 for all edges. In order to correctly incorporate the functional similarities in the analysis, you have to copy the rfunsim attribute to the weight attribute. To do so, select all edges by pressing Alt+a and then use Cytoscape's Attribute Batch Editor, which you find in the Edge Attribute Browser as shown in Fig. 4. Now you should open the NetworkPrioritizer preferences via NetworkPrioritizer -> Preferences as shown in Fig. 5. Here, you adjust the Graph-Related settings to the values corresponding to Fig. 6: weight should be used as weight during the network analysis. The distance should be set to Tuned Inverted Distance and the weights should contribute by alpha = 0.8. All centralities should be computed per connected component. When the preferences are adjusted, click Apply and close the dialog.

Fig. 4: Cytoscape's Attribute Batch Editor. Fig. 5: How to find the NetworkPrioritizer preferences. Fig. 6: The values of the preferences for the analysis.

To start the computations, choose NetworkPrioritizer -> Rank nodes -> According to centralities to be computed and select all measures in the upcoming dialog, as is shown in Fig. 7 and 8, respectively. After a couple of seconds, the Ranking Manager should pop up and show the resulting primary rankings (see Fig. 9).

Fig. 7: How to start the computation of centralities. Fig. 8: Selection of centrality measures to be computed. Fig. 9: Ranking Manager showing the primary rankings.

To obtain the final ranking, click on Aggregate Rankings and select all available primary rankings (see Fig. 10). The Ranking Manager should then show the aggregated ranking under the name that was entered in the Ranking Selection dialog, as shown in Fig. 11.

Fig. 10: Ranking Selection dialog. Fig. 11: Ranking Manager showing the aggregated ranking.