Superpathways plugin for Cytoscape

Superpathways: Tutorial and usage

1. Introduction

Biological pathways are highly interconnected. To find out how different processes influence each other, it can be useful to visualize the connectivity between pathways. Superpathways is a Cytoscape plugin (Cytoscape 2.6) dedicated to providing useful visualization and an analysis tool for such purpose. It allows the user to search and select multiple pathways from Wiki Pathways, and then to merge them into a single network. Nodes that represent the same biological entity (e.g. gene, protein or molecule) would be merged. A visualization that indicated the origin pathway of each node is provided.

1.1 Structure of this Document

Section 1. Information on getting started with the plugin and tutorial

Section 2. Tutorial for searching and selecting pathways from Wiki Pathways.

Section 3. Tutorial for using Superpathways to analyze the connectivity between pathways

1.2 Loading the Superpathways Plugin

Before starting the tutorial, please make sure that you have downloaded and installed a current version of Cytoscape (2.6 or higher) on your computer. Next, generate the Superpathways plugin superpathways.jar according to the instruction in the Source Code section. Put the jar file, into the folder called plugins in the Cytoscape directory on your computer. The directory on a Windows computer is shown in Figure 1.

Figure 1 Location of the plugins folder in the Cytoscape application

2. Tutorial for Searching/ Selecting Pathways from Wiki Pathways

2.1

Complete the instruction in section 1.2 above before starting the tutorial to make sure that the plugin are properly installed.

2.2

Open the Superpathways plugin by selecting Plugins->Superpathways on the top Cytoscape menu, and then a user interface window would pop up (see Figure 2).

Figure 2 Location of Superpathways plugin in Cytoscape Window and its UI

2.3

The major UI panel allows the user to search pathways from Wiki Pathways by inputting pathway name (e.g. Apoptosis), gene or protein name (e.g. p53), or any page content (cancer). You can also choose the type of organism to decrease the range of searching. After clicking ‘Search’ button, a list of candidate pathways’ information appears in the result table, including pathway name, ID, and organism to which they belong. The user can add one interesting pathway or several of them (using ‘Ctrl’) to the ‘Selected pathways’ list by clicking the button ‘Add’. When the user is not quite sure about which pathways he's going to merge, the tool also provides help with 'Find Related Pathways' option. We'll introduce this part in the following 2.4 section.

If you want to add more pathways of interest, and cannot find it in the current result table, you can search them by inputting different keywords. 'Remove' button allows the user to delete one or several previous selected pathways. When you perform a totally new task (merging a new set of interesting pathways), it is better to click “Clear” button to empty the ‘Selected Pathways’ blocks.

2.4

The help dialog offers the user help when he is not clear about which pathways might be interconnected with the ones of interest. In such case, he first needs to choose the pathway of interest in the result table, and then click the ‘Find Related Pathways’ button in the major panel, the help dialog would pop up. In this dialog, the user needs to indicate the range of sharing nodes number. E.g., if you want to figure out which pathways share between 3 and 20 nodes with ‘p53 signal pathway’ (suppose it is already chosen in the result table), then he can set the ‘Sharing Minimum Nodes’ as 3, and the ‘Maximum Nodes’ as 20 (see Figure 3). After clicking the ‘Search’ button on the dialog, the tool would return the information (including Pathway Name, ID, and number of shared nodes) of all the candidate pathways each sharing between 3 and 20 nodes with ‘p53 signal pathway’. The user can add one or several of them (using ‘Ctrl’) to the ‘Selected Pathways’ by clicking the button ‘Add’.

Figure 3 an example to show how to use ‘Search Help’ tab

3. Tutorial for using Superpathways to analyze the connectivity between pathways

3.1.

For the purpose of prediction and better understanding of the relationship between pathways, the tool provides the user with an option to see whether all the selected pathways have interconnectivity. ‘Common Node View’ is proposed here for this purpose. The user can obtain this view for all the pathways in the ‘Selected Pathways’ block by clicking the ‘Common Node View’ button on the bottom of the ‘Search/Select’ tab. In the view, each node is used to represent a pathway, and node label is the combination of pathway name and its id. The number of sharing nodes between every two pathways is shown on a line connecting their two corresponding nodes. Through this view, the user can easily figure out whether the two selected pathways share any node. If not, then there is no need to merge them.

For instance, you can find an example in Figure 4. Pathway ‘Apoptosis Mechanisms (WP168)’ shares four nodes with pathway ‘G1 to S cell cycle control (WP413)’, while pathways ‘MAPK signaling pathway (WP382)’ and ‘Apoptosis (WP254)’ have 21 nodes in common.

Figure 4 Common Node View

It is worth noticing that in common node view we apply Force Directed layout (with weight) by default. The number of shared nodes is used as weight. Therefore, pathways sharing many nodes would be close together, and then the user can immediately figure out which pathways are more related than others. The closer the two pathways are, the more nodes they share between them. Therefore, the two later pathways ‘MAPK signaling pathway (WP382)’ and ‘Apoptosis (WP254)’ are more interconnected than the two former ones.

Another useful option is that the attribute ‘weight’ (marked in the red ellipse) storing the number of shared nodes is numeric, thus the user can apply his favorite weighted layout manually after the network is generated.

3.2

After looking at the Common Node View to make sure whether the selected pathways are interconnected, the user can click the ‘Merge’ button on the bottom of the ‘Search/Select’ tab to get the merged network for the pathways. He can indicate the original pathways of each node by two means. One available way is from the color of node or pie, and we’re explaining how to figure out the original pathways through an example (see Figure 5).

(a)

(b)

Figure 5 an example to show how to figure out the original pathways of a node or pie in the merged network view

Figure 5.a is the screenshot of the merged network view of four pathways (Apoptosis (WP254), Apoptosis Modulation by HSP70 (WP384), p38 MAPK Signaling Pathway (BioCarta?) (WP400), and MAPK Cascade (WP422)). You can find that in the VizMapper? tab of the Control Panel we assign a color to each pathway, and you can use this to figure out the original pathway for each node and pie. E.g., pathway Apoptosis is assigned a red color; therefore all the red nodes in the merged network view belong to this pathway. Pies are used to indicate that the nodes they represent are shared by several pathways. The colors that compose a pie can tell you which pathways share the corresponding entity. E.g., the pie in the left green circle consist of four colors, then the entity represented by this pie is shared by all four pathways; whereas the one in the right green circle consist of three colors (purple, light blue, and red), then it means the corresponding entity is shared by three pathways including Apoptosis, MAPK Cascade, and p38 MAPK Signaling Pathway (BioCarta?).

It is worth noticing that there is also a correspondence of colors in the common node view (See Figure 5.b) and the merged network, i.e. the color of the node representing a pathway in the Common Node View is the same as the one that we assign to the pathway in the VizMapper?. Therefore, the user can use this correspondence to tell the original pathways of each node and pie as well.

The other way that the user can tell the origin of each node is from the node attribute ‘Source Pathway’. In Figure 5.a, the red circle right below the title of ‘Data Panel’ shows the attribute list icon. To see the ‘Source Pathway’ attribute in the data panel, you need to left-click on this icon to get a list of all attributes, then check the boxes for the attribute ‘Source Pathway’. Through this attribute, it is easy for the user to figure out the original pathways of a node or pie.

3.3

When clicking the ‘Merge’ button, unmerged pathways are also visualized and colored by default. The rule for coloring is to color the nodes that are unique for a pathway, i.e. the nodes that are shared with other selected pathways would not be colored. In this way, the user can easily know which nodes are common ones and which are not by just looking at unmerged pathways view. Also, we provide the option to the user to switch off the coloring if it is done in a way he doesn’t like. Figure 6.a shows the Visual Style with the name of ‘GPML-extension’ (marked in red circle) that we use for coloring unmerged pathways. For turning off the coloring, the user only needs to change to the Visual Style to ‘GPML’ (marked in red circle) in Figure 6.b.

Figure 6(a) GPML-extension Visual Style can show the coloring of unmerge pathways

Figure 6(b) GPML Visual Style is the one without coloring for unmerged pathway

3.4

Currently, the tool applies the Grid Layout by default. Later, the user needs to change it to other appropriate one manually. Some of the existing Cytoscape layouts already provide a great toolkit for investigating the merged pathways. Which layout algorithm should be applied later really depends on what you want to see. If you want to see which nodes are shared and which are not, then Group Attribute Layout (see Figure 7.a) would be perfect. If you want to see which reactions or paths are common between the pathways, then Force-Directed (see Figure 7.b) or Organic layouts would be great. Therefore, we leave this to the user, and he can choose a proper one to meet his need.

(a) Group Attribute Layout

(b) Force-Directed

Figure 7 two useful layouts for the merged pathway view

Source code

The source code for this plugin can be found in our subversion repository:  http://svn.bigcat.unimaas.nl/pathvisio/trunk/tools/superpathways

To build the superpathways jar file, run:

ant jar -Dcytoscape.dir=/path/to/cytoscape