BioNetBuilder Tutorial

The following tutorial will familiarize you with the BioNetBuilder PlugIn. It assumes you already know how to start Cytoscape with the PlugIn loaded (either through the BioNetBuilder Java Web-Start, or by including the PlugIn jar in the Cytoscape plugIns directory, consult the BioNetBuilder home page to learn how to start the plugIn using this method).

1. Start the PlugIn
Under the PlugIns menu there are two BioNetBuilder menu items. "Set BioNetBuilder Connection..." allows you to set the URL location of the BioNetBuilder server. We provide a default server, but if it is used by many people at the same time, it may be slow. We provide instructions on how to set up your own server here. "BioNetBuilder Wizard..." provides a wizard to create your network, click on it to start.


2. Select a species
Type in the text field the species for your network. The name of the species can be incomplete. Letter case is ignored. Then press "Search". Species that match your entered string will be displayed. Select ONE. You can select a different species for each network you create.


3. Select nodes for your network
The next step allows you to specify the nodes in your network. There are 4 different ways of specifying network nodes, and you can select (using the check boxes on the left) the ones you wish to use. If you skip this step, then the nodes in your network will be specified in the next wizard step, where you select interaction sources.


"Nodes with selected annotations..."
You can select nodes with certain Gene Ontology annotations. For example, you can select all nodes that are localized in the nucleus AND whose biological function is unknown. Because you want the genes to have BOTH annotations (nucleus AND unknown function) select the "Use AND boolean operator" option. If you want to get all nodes that have at least ONE of the annotations (nucleus OR unknown function), then don't select it. You can select as many annotations as you want. The genes are found recursively: all genes that have an annotation within a selected annotation tree are searched. Press the "OK" button to get the annotated genes in your network.

<
"Nodes from my list..."
You can load a .txt file that contains a list of nodes you want to use for your network. The accepted IDs for genes are:
Each line in the file corresponds to one gene, and it should contain a prefix indicating the type of ID used for that gene ("GI:","RefSeq:",or,"ORF:").


"Nodes from NCBI Taxonomy..."
You can enter one or more (comma separated) common name strings for genes and search for their GI RefSeq amino acid numbers. You can then select the matches that you wish to include in your network.


"Nodes from loaded networks..."
If Cytoscape contains networks you already created using the BioNetBuilder, then you can improve performance by using nodes in these networks for your new network (instead of retrieving the nodes from the server again). If you use this option, you can select which networks to use, and also, whether or not to use selected nodes in these networks.

Pay attention to the "Num Nodes" column! The "Num Nodes" column indicates how many nodes are going to be included in your network for each selected method. If you selected a method, and this number is zero, that means that no nodes are being included using this method. This could be because: there are no genes annotated with the selected GO annotations, none of the gene IDs in your text file are accepted ID types or exist in the database, or, you forgot to select matches in the common name search tool.


4. Select interaction sources
This step allows you to select the interaction databases you wish to use for your network, set parameters for each database, and calculate the number of interactions your network will have given these settings.


To select an interaction database, select its check box to the left. To set database parameters, press its button (for example, "KEGG...").
If you selected a set of nodes for your network in the previous step, then this step will find connecting interactions among them. You can also connect to them their first neighbors. To do this, select the "Add first neighbors of nodes" option. This option is only available if you selected a set of nodes in the previous step.
If you didn't select a set of nodes for yout network in the previous step, then ALL interactions from the selected databases will be in your network. Note that this may result in very large networks that may crash the program. Try to create networks with less than 100,000 edges. You can calculate the number of edges in your network by pressing the "Calculate number of edges from selected dastabases" button.


The HPRD database has a license that prevents if from being used by for profit institutions. If you wish to use it, send us an email to request for the user name and password.

5. Set node label priorities
This step allows you to choose how you want the nodes in your network to be visually labeled. In the example below, all genes should be labeled with their official gene name, but, if there is no gene name for a node, then it will be labeled with it's ORF, but, if there is no ORF (for yeast there is), then it will be labeled with its product name, and so forth. The last label priority is the original ID used in the database which could only be meaningful in that database.


6. Select node attributes
In this step you can choose from a set of attributes to attach to the nodes in your network. The most useful ones are the URL attributes since they allow you to directly connect a node to a web page with information about that node in a public database like NCBI, PIR, Prolinks, The Yeast Resource Center (called "HPFP_URL") etc. You will learn how to invoke these URLs from Cytoscape later in this tutorial.


7. Name your network and make it look pretty
Simply name your network and press the "Finish" button. BioNetBuilder will then retrieve nodes, edges, and attributes from the server, and will build a Cytoscape network. If you enter the name of an existing network that is loaded into Cytoscape, then, the new nodes and edges will be added to this network.



If all goes well, you will see a network with the nodes set in rows and columns as shown below.


You can then layout the graph ("Layout->yFiles->Organic" is my favorite), and get a more visually pleasing hairball:


As you can see in the image above, edges have different colors. They are color coded according to their interaction type, which is stored in the interaction databases you selected. To set edge colors, click on the pie chart icon in the Cytoscape toolbar, and then press the "Define" button in the "VizMapper" dialog. Then press the "Edge Attributes" button, and select the "Edge Color" tab. You can then change colors for your edges by clicking on the color for each interaction type. Press the "Apply to Network" button when you are done.


You now have a network that you can analyse and study using all of the capabilities Cytoscape has. You can save this network for later sessions using available Cytoscape tools which are clearly explained in the Cytoscape manual (see Cytoscape home).

How to launch a web browser with attached node URLs
Select a set of nodes (or select them all) by using your mouse, or other Cytoscape node selection methods available under the "Select" menu.
Then, for easy viewing, pop-up the attributes browser (located at the bottom of the Cytoscape Desktop by default, also called "CytoPanel 2"). To pop it out of the desktop, click on the icon located on its top right corner. You can then resize the attributes browser so that it is larger (by using your mouse and draging one of its edges).
Now, click on the blue arrow icon to select the attributes you wish to display. Select all the ones that end with "_URL".


When you click on a http entry in the attributes browser, you should see a web browser pop-up with that URL loaded:


Go back to BioNetBuilder home
Last updated: 7.13.2006