This document is licensed under the Creative Commons license, 2006

Authors: The Cytoscape Collaboration

The Cytoscape project is an ongoing collaboration between:

Funding for Cytoscape is provided by a federal grant from the U.S. National Institute of General Medical Sciences (NIGMS) of the National Institutes of Health (NIH) under award number GM070743-01. Corporate funding is provided through a contract from Unilever PLC.

Cytoscape is a project dedicated to building open-source network visualization and analysis software. A software "Core" provides basic functionality to layout and query the network and to visually integrate the network with state data. The Core is extensible through a plug-in architecture, allowing rapid development of additional computational analyses and features.

Cytoscape's roots are in Systems Biology, where it is used for integrating biomolecular interaction networks with high-throughput expression data and other molecular state information. Although applicable to any system of molecular components and interactions, Cytoscape is most powerful when used in conjunction with large databases of protein-protein, protein-DNA, and genetic interactions that are increasingly available for humans and model organisms. Cytoscape allows the visual integration of the network with expression profiles, phenotypes, and other molecular state information, and links the network to databases of functional annotations.

The central organizing metaphor of Cytoscape is a network (graph), with genes, proteins, and molecules represented as nodes and interactions represented as links, i.e. edges, between nodes.

Development

Cytoscape is a collaborative project between the Institute for Systems Biology (Leroy Hood lab), the University of California San Diego (Trey Ideker lab), Memorial Sloan-Kettering Cancer Center (Chris Sander lab), the Institut Pasteur (Benno Schwikowski lab), Agilent Technologies (Annette Adler lab) and the University of California, San Francisco (Bruce Conklin lab).

Visit http://www.cytoscape.org for more information.

License

Cytoscape is protected under the GNU LGPL (Lesser General Public License). The License is included as an appendix to this manual, but can also be found online: http://www.gnu.org/copyleft/lesser.txt. Cytoscape also includes a number of other open source libraries, which are detailed in theCytoscape User Manual/Acknowledgements below.

What’s New in @version@

Cytoscape version @version@ contains several new features, plus improvements to the performance and usability of the software. These include:

• Web Service Client Manager framework to integrate web service clients into Cytoscape.

• Web Service client plugins for downloading networks from PathwayCommons, IntAct, and NCBI Entrez Gene.

• Annotation import web service plugin for BioMart. This is mainly for ID translation/synonym mapping.

• Cytoscape themes.
• Dynamic filters.
• Network Manager supports multiple network selection.
• Label Positioning has been improved.
• Session saving occurs in memory.
• XGMML Improvements.
• Network loading improvements.
• Linkout integrated with attribute browser.
• Extra sample Visual Styles using new visual properties introduced in 2.5.
• Many, many bug fixes!

Cytoscape is a Java application verified to run on Linux, Windows, and Mac OS X. Although not officially supported, other UNIX platforms such as Solaris or FreeBSD may run Cytoscape if Java version 5 or later is available for the platform.

System requirements

The system requirements for Cytoscape depend on the size of the networks the user wants to load, view and manipulate.

Getting Started

Install Java

If not already installed on your computer, download and install Java SE 5 or 6. Cytoscape 2.5 will no longer run with Java version 1.4.x or lower. You must install Java SE 5 or 6!!!

These can be found at:

Java SE 5 Java SE 6

In general, Java SE 6 is faster than 5. If your machine is compatible with the 6 series, please try version 6.

Install Cytoscape

There are a number of options for downloading and installing Cytoscape. All options can be downloaded from the http://cytoscape.org website.

• Automatic installation packages exist for Windows, Mac OS X, and Linux platforms.
• You can install Cytoscape from a compressed archive distribution.
• You can build Cytoscape from the source code.

• You can check out the latest and greatest software from our Subversion repository.

Cytoscape installations (regardless of platform) containing the following files and directories:

* From Ideker et al., Science 292:929 (2001)

** Obtained from data hosted at http://www.blueprint.org/bind/bind_downloads.html

** From von Mering et al., Nature, 417:399 (2002) and Lee et al, Science 298:799 (2002)

**** Created from Cytoscape tutorial web page. Original data sets are available at: http://www.cytoscape.org/cgi-bin/moin.cgi/Data_Sets/ from "A merged human interactome" by Andrew Garrow, Yeyejide Adeleye and Guy Warner (Unilever, Safety and Environmental Assurance Center).

Launch the application

Double-click on the icon created by the installer or by running cytoscape.sh from the command line (Linux or Mac OS X) or double-clicking cytoscape.bat (Windows). Alternatively, you can pass the .jar file to Java directly using the command java -Xmx512M -jar cytoscape.jar -p plugins. The -Xmx512M flag tells java to allocate more memory for Cytoscape and the -p plugins option tells cytoscape to load all of the plugins in the plugins directory. Loading the plugins is important because many key features like layouts, filters and the attribute browser are included with Cytoscape as plugins in the plugins directory. See the Command Line chapter for more detail. In Windows, it is also possible to directly double-click the .jar file to launch it. However, this does not allow specification of command-line arguments (such as the location of the plugin directory).

When you succeed in launching Cytoscape, a window will appear that looks like this (captured on Mac OS 10.4):

Note on Memory Consumption

For users interested in loading large networks, the amount of memory needed by Cytoscape will increase. Memory usage depends on both number of network objects (nodes+edges) and the number of attributes. Here are some rough suggestions for memory allocation:

Suggested Memory Size Without View

Suggested Memory Size With View

Overall Memory Size for Cytoscape

To increase the maximum memory size for Cytoscape, you can specify it from command line. For example, if you want to assign 1GB of memory, type:

java -Xmx1GB -jar cytoscape.jar -p plugins

from the command line.

Stack Size

There is one more option related to memory allocation. Some of the functions in Cytoscape use larger stack space (a temporary memory for some operations, such as Layout). Since this value is set independently from the Xmx value above, sometimes Layout algorithms fails due to the out of memory error. To avoid this, you can set larger heap size by -Xss option. If layout fails for large networks, please try the following:

java -Xmx1GB -Xss10M -jar cytoscape.jar -p plugins

The option -Xss10M means set the heap size to 10MB. In many cases, this solves out of memory error for Layouts.

Randomly generated scale-free network with 500K nodes and 500k edges: If memory parameters are set properly, you can visualize huge networks. In this example, about 5GB of memory is used by Cytoscape. Stack size is set to 10MB. To use large memory space (4GB+) you need 64-bit version operating system AND 64-bit version Java SE 5 or 6.

Note: Some of the web service clients are multi-thread programs and each thread uses the memory size specified by -Xss option. If web service clients fails due to the out of memory error, please reduce the stack size and try again.

For more details, see How to increase memory for Cytoscape.

Note on Directory Location

For the application to work properly, all files should be left in the directory in which they were unpacked. The core Cytoscape application assumes this directory structure when looking for the various libraries needed to run the application. If you are adventurous, you can get creative with the $CLASSPATH and/or the cytoscape.jar manifest file and run Cytoscape from any location you want.

When a network is loaded, Cytoscape will look something like the image below:

The main window here has several components:

• The menu bar at the top (see below for more information about each menu).
• The toolbar, which contains icons for commonly used functions. These functions are also available via the menus. Hover the mouse pointer over an icon and wait momentarily for a description to appear as a tooltip.
• The network management panel (top left panel). This contains an optional network overview pane (shown at the bottom left).
• The main network view window, which displays the network.
• The attribute browser panel (bottom panel), which displays attributes of selected nodes and edges and enables you to modify the values of attributes.

The network management and attribute browser panels are dockable tabbed panels known as CytoPanels. You can undock any of these panels by clicking on the Float Window control in the upper-right corner of the CytoPanel.

If you select this control, e.g. on the attribute browser panel, you will now have two Cytoscape windows, the main window, and a new window labeled CytoPanel 2, similar to the one shown below. Popup will be displayed when you put the mouse pointer on a cell.

Note that CytoPanel 2 now has a Dock Window control. If you select this control, the window will dock onto the main window.

Cytoscape also has an editor that enables you to build and modify networks interactively by dragging and dropping nodes and edges from a palette onto the main network view window. The Node shapes and Edge arrows on the palette are defined by the currently used Visual Style. To edit a network, just select the Editor tab on CytoPanel 1. An example of an editor, with the palette contained in CytoPanel 1 and defined by the BioMoleculeEditor Visual Style, is shown below.

The Menus

File

The File menu contains most basic file functionality: File → Open for opening a Cytoscape session file; File → New for creating a new network, either blank for editing, or from an existing network; File → Save for saving a session file; File → Import for importing data such as networks and attributes; and File → Export for exporting data and images. Also, File → Print allows printing, while File → Quit closes all windows of Cytoscape and exits the program.

Edit

The Edit menu contains Undo and Redo functions which undo and redo edits made in the Attribute Browser, the Network Editor and to layout.

There are also options for creating and destroying views (graphical representations of a network) and networks (the raw network data – not yet visualized), as well as an option for deleting selected nodes and edges from the current network. All deleted nodes and edges can be restored to the network via Edit → Undo. Editing preferences for properties and plugins is found under Edit → Preferences → Properties... .

View

The View menu allows you to display or hide the network management panel (CytoPanel 1), the attribute browser (CytoPanel 2), the Network Overview (in CytoPanel 1), and the VizMapper.

Select

The Select menu contains different options for selecting nodes and edges. It also contains the Select → Use Filters option, which allows filters to be created for automatic selection of portions of a network whose node or edge attributes meet a filtering criterion.

Layout

The Layout menu has an array of features for visually organizing the network. The features in the top portion of the network (Rotate, Scale, Align and Distribute) are tools for manipulating the network visualization. The bottom section of the menu lists a variety of layout algorithms which automatically lay a network out.

Plugins

The Plugins menu contains options for managing (install/update/delete) your plugins and may have options added by plugins that have been installed, such as the Agilent Literature Search or Merge Networks. Depending on which plugins are loaded, the plugins that you see may be different than what appear here.

Help

The Help menu allows you to launch the online help viewer and browse the table of contents for this manual. The “About…” option displays information about the running version of Cytoscape.

Network Management

Cytoscape 2.3 and newer versions allow multiple networks to be loaded at a time, either with or without a view. A network stores all the nodes and edges that are loaded by the user and a view displays them. You can have many views of the same network. Networks (and their optionally associated views) can be organized hierarchically.

An example where a number of networks have been loaded and arranged hierarchically is shown below:

The network manager (top-right tree view in CytoPanel 1) shows the networks that are loaded. Clicking on a network here will make that view active in the main window, if the view exists (green highlighted networks only). Each network has a name and size (number of nodes and edges), which are shown in the network manager. If a network is loaded from a file, the network name is the name of the file.

Some networks are very large (thousands of nodes and edges) and can take a long time to display. For this reason, a network in Cytoscape may not contain a ‘view’. Networks that have a view are highlighted in green and networks that don’t have a view are highlighted in red. You can create or destroy a view for a network by right-clicking the network name in the network manager or by choosing the appropriate option in the Edit menu. You can also destroy previously loaded networks this way. In the picture above, seven networks are loaded, six green ones with views and one red one without a view.

Certain operations in Cytoscape will create new networks. If a new network is created from an old network, for example by selecting a set of nodes in one network and copying these nodes to a new network (via the File → New → Network option), it will be shown as a child of the network that it was derived from. In this way, the relationships between networks that are loaded in Cytoscape can be seen at a glance. Networks in the top part of the tree in the figure above were generated in this manner.

The available network views are also arranged as multiple overlapping windows in the network view window. You can maximize, minimize, and destroy network views by using the normal window controls for your operating system.

Arrange Network Windows

When you work on multiple networks, you can arrange the network view windows from View → Arrange Network Windows. You can re-arrange the network location by these commands.

The Network Overview Window

The network overview window shows an overview (or ‘bird’s eye view’) of the network. It can be used to navigate around a large network view. The blue rectangle indicates the portion of the network currently displayed in the network view window, and it can be dragged with the mouse to view other portions of the network. Zooming in will cause the rectangle to appear smaller and vice versa.

Cytoscape recognizes a number of optional command line arguments, including run-time specification of network files, attribute files, and session files. This is the output generated when the cytoscape is executed with the "-h" or "--help" flag:

usage: java -Xmx512M -jar cytoscape.jar [OPTIONS]
 -h,--help                      Print this message.
 -v,--version                 Print the version number.
 -s,--session <file>        Load a cytoscape session (.cys) file.
 -N,--network <file>     Load a network file (any format).
 -e,--edge-attrs <file>    Load an edge attributes file (edge attribute format).
 -n,--node-attrs <file>   Load a node attributes file (node attribute format).
 -m,--matrix <file>        Load a node attribute matrix file (table).
 -p,--plugin <file>         Load a plugin jar file, directory of jar files,
                                     plugin class name, or plugin jar URL.
 -P,--props <file>         Load cytoscape properties file (Java properties
                                    format) or individual property: -P name=value.
 -V,--vizmap <file>      Load vizmap properties file (Java properties format).

Any file specified for an option may be specified as either a path or as a URL. For example you can specify a network as a file (assuming that myNet.sif exists in the current working directory): cytoscape.sh -N myNet.sif. Or you can specify a network as a URL: cytoscape.sh -N http://example.com/myNet.sif.

All options described above (including plugins) can be loaded from the GUI once Cytoscape is running.

Managing Properties

The Cytoscape Properties editor, accessed via Edit → Preferences → Properties…, is used to specify general and default properties. Properties are now stored in Cytoscape session files, so changes to general properties will be saved as part of the current session, but will only carry over to subsequent sessions if they are set as defaults or exported using the File → Export function.

Cytoscape properties are configurable via Add, Modify and Delete operations.

Some common properties are described below.

Setting Default Properties

It is possible to alter the default properties for Cytoscape.

Edit the properties via Edit → Preferences → Properties... and check the Make Current Cytoscape Properties Default checkbox. This will save the current properties to the .cytoscape directory, where they will then be applied to all of your Cytoscape sessions from that point on. Otherwise, Cytoscape will automatically save the properties used in a particular session inside its .cys session file, while the default properties will be applied at the beginning of subsequent sessions.

Managing Bookmarks

Cytoscape contains a pre-defined list of bookmarks, which point to sample network files located on the Cytoscape web server. Users may add, modify, and delete bookmarks through the Bookmark manager, accessed by going to Edit → Preferences → Bookmarks… .

There are currently two types of bookmarks: network and annotation. Network bookmarks are URLs pointing to network files available on the Internet. These are nomal networks that can be loaded into Cytoscape. The annotation bookmarks are URLs pointing to ontology annotation files. The annotation bookmarks are only used when importing an ontology.

Managing Proxy Servers

You can define and configure a proxy server for Cytoscape by going to Edit → Preferences → Proxies… .

After the proxy server is set, all network traffic related to loading a network from URL will pass through the proxy server. Other plugins use this capability as well.

There are 4 different ways of creating networks in Cytoscape:

1. Importing pre-existing, formatted network files.
2. Importing pre-existing, unformatted text or Excel files.
3. Importing networks from Web Service.
4. Creating an empty network and manually adding nodes and edges.

Import Fixed-Format Network Files

Network files can be specified in any of the formats described in the Supported Network Formats chapter. Networks are imported into Cytoscape through the "Import Network" window, which can be accessed by going to File → Import → Network (multiple file types). The network file can either be located directly on the local computer, or found on a remote computer (in which case it will be referenced with a URL).

Load Networks from Local Computer

By default, Cytoscape loads networks from the local computer.

The Import Networks dialog shows a default setting of "Data Source Type: Local," meaning that network files from the local computer will be available for importing. Choose the correct file by clicking on the Select button (only file types that Cytoscape recognizes will be shown), and then load the network by clicking on the Import button. Some sample network files of different types have been included in the sampleData folder in Cytoscape.

Network files in SIF, GML, and XGMML formats may also be loaded directly from the command line using the –N option.

Load Networks from a Remote Computer (URL import)

The Import Networks dialog is also capable of importing network files using a URL. To do this, set the Data Source Type to Remote and insert the appropriate URL, either manually or using URL bookmarks. Bookmarked URLs can be accessed by clicking on the arrow to the right of the text field (see the Bookmark Manager in Preferences for more details on bookmarks). Also, you can drag and drop links from web browser to the URL text box. Once a URL has been specified, click on the Import button to load the network.

Importing networks from URL addresses has an important caveat. Because Cytoscape determines file type primarily (not exclusively) by file extension, it can have trouble importing networks with URLs that don't end in a human readable file name. If Cytoscape does not recognize a meaningful file name and extension in the URL, it will attempt to guess the type of file based on MIME type. If the MIME type is not recognizable to any of our import handlers, then the import will fail.

Another issue for network import is the presence of firewalls, which can affect which files are accessible to a computer. To work around this problem, Cytoscape supports the use of proxy servers. To configure the proxy server, go to Edit → Preferences→ Proxy Server... . This is further described in the Preferences chapter.

Import Free-Format Table Files

Introduced in version 2.4, Cytoscape now supports the import of networks from delimited text files and Excel workbooks using Edit → Import → Network from Table (Text/MS Excel)... . An interactive GUI allows users to specify parsing options for specified files. The screen provides a preview that shows how the file will be parsed given the current configuration. As the configuration changes, the preview updates automatically. In addition to specifying how the file will be parsed, the user must also choose the columns that represent the Source nodes, the Target nodes, and an optional edge interaction type.

Supported Files

The "Import Network from Table" function supports delimited text files and single-sheet Microsoft Excel Workbooks. The following is a sample table file:

source  target  interaction  boolean attribute  string attribute        floating point attribute
YJR022W YNR053C pp      TRUE    abcd12371       1.2344543
YER116C YDL013W pp      TRUE    abcd12372       1.2344543
YNL307C YAL038W pp      FALSE   abcd12373       1.2344543
YNL216W YCR012W pd      TRUE    abcd12374       1.2344543
YNL216W YGR254W pd      TRUE    abcd12375       1.2344543

The network table files should contain at least two columns: source nodes and target nodes. The interaction type is optional in this format. Therefore, a minimal network table looks like the following:

YJR022W YNR053C
YER116C YDL013W
YNL307C YAL038W
YNL216W YCR012W
YNL216W YGR254W

One row in a network table file represents an edge and its edge attributes. This means that a network file is considered a combination of network data and edge attributes. A table may contain columns that aren't meant to be edge attributes. In this case, you can choose not to import those columns by clicking on the column header in the preview window. This function is useful when importing a data table like the following (1):

Unique ID A     Unique ID B     Alternative ID A        Alternative ID B        Aliases A       Aliases B       Interaction detection methods   First author surnames   Pubmed IDs      species A       species B       Interactor types        Source database Interaction ID  Interaction labels      Cross-references        Associated Files        Experiment files        Experiment labels       Different techniques    Different Pubmed articles       Different sources       Weight

7205    5747    TRIP6   PTK2    Q15654  Q05397-1        vv|HPRD Currently not available 14688263|15892868(Marcotte)     Mammalia        Homo sapiens    protein|protein HPRD|Marcotte   0       Thyroid hormone receptor interactor 6-FAK-|PTK2-TRIP6   NA(HPRD)|NA(Marcotte)   HPRD/02859_psimi.xml|other/ORIGINAL_DATA_MARCOTTE.txt   vv(HPRD/02859_psimi.xml)|HPRD(other/ORIGINAL_DATA_MARCOTTE.txt) 17651(ExptRef)|Marcotte 2       2       2       2

4174    7311    MCM5    UBA52   P33992  P62987  neighbouring_reaction   Currently not available 15608231(Reactome)      Homo sapiens    Homo sapiens    protein|protein Reactome        1       P33992-P62988   Reaction:68944<->Reaction:68946(Reactome)|Reaction:68946<->Reaction:68944(Reactome)     other/ORIGINAL_DATA_MARCOTTE.txt        neighbouring_reaction(other/REACTOMEhomo_sapiens.interactions.txt)      Reactome        1       1       1       1

7040    7040    TGFB1   TGFB1   P01137  P01137  nmr: nuclear magnetic resonance Currently not available 8679613 Homo sapiens    Homo sapiens    protein|protein BIND    2       TGFB1-TGFB1-    72085(BIND)     BIND/bind_taxid9606.1.psi.xml   nmr: nuclear magnetic resonance(BIND/bind_taxid9606.1.psi.xml)  NotAvailable    1       1       1       1

This data file is a tab-delimited text and contains network data (interactions), edge attributes, and node attributes. To import network and edge attributes from this table, you need to choose Unique ID A as source, Unique ID B as target, and Interactor types as interaction type. Then you need to turn off columns used for node attributes (Alternative ID A, species B, etc.). Other columns can be imported as edge attributes.

The network import function cannot import node attributes - only edge attributes. To import node attributes from this table, please see the Attributes section of this manual.

Note (1): This data is taken from the A merged human interactome datasets by Andrew Garrow, Yeyejide Adeleye and Guy Warner (Unilever, Safety and Environmental Assurance Center, 12 October 2006). Actual data files are available at http://www.cytoscape.org/cgi-bin/moin.cgi/Data_Sets/

Basic Operations

To import network text/Excel tables, please follow these steps:

1. Select File → Import → Network from Table (Text/MS Excel)...

2. Select a table file by clicking on the Select File button.
3. Define the interaction parameters by specifying which columns of data contain the Source Interaction, Target Interaction, and Interaction Type. Setting the Interaction Type as Default Interaction will result in all interactions being given the value pp; this value can be modified in Advanced Options (below).
4. (Optional) Define edge attribute columns, if applicable. Network table files can have edge attribute columns in addition to network data.

   • Enable/Disable Attribute Column - By left-clicking on a column header in the preview table, you can enable/disable edge attributes. If the header is checked and entries are blue, the column will be imported as an edge attribute. For example, the table below shows that columns 1 through 3 will be used as network data, column 4 will not be imported, and columns 5 and 6 will be imported as edge attributes.

     • 

   • Change Attribute Name and Data Types - If you right-click on a column header in the preview table, you can modify the attribute name and data type. For more detail, see "Modify Attribute Name/Type" below.

5. Click the Import button.

Import List of Nodes Without Edges

Table Import feature supports list of nodes without edges. If you select source column only, it creates a network without interactions. This feature is useful with node expansion function available from some web service clients. Please read the section Importing Networks from External Database for more detail.

Advanced Options

You can select several options by checking the Show Text File Import Options checkbox.

• Delimiter: You can select multiple delimiters for text tables. By default, Tab and Space are selected as delimiters.
• Preview Options: When you select a network table file, the first 100 entries will be displayed in the Preview panel. To display more entries, change the value for this option. If you want to show all entries in the file, select "Show all entries in the file". You will need to click the Reload button to update the Preview panel.
• Attribute Names
  • Transfer first line as attribute names: Selecting this option will cause all edge attribute columns to be named according to the first data entry in that column.
  • Start Import Row: Set which row of the table to begin importing data from. For example, if you want to skip the first 3 rows in the file, set 4 for this option.
  • Comment Line: Rows starting with this character will not be imported. This option can be used to skip comment lines in text files.
• Network Import Options: If the Interaction Type is set to Default Interaction, the value here will be used as the interaction type for all edges.

Modify Attribute Name/Type

Attribute names and data types can be modified here.

• Modify Attribute Name - just enter a new attribute name and click OK.
• Modify Attribute Data Type - The following attribute data types are supported:
  • String
  • Boolean (True/False)
  • Integer
  • Floating Point
  • List of (one of) String/Boolean/Integer/Floating Point

Cytoscape has a basic data type detection function that automatically suggests the attribute data type of a column according to its entries. This can be overridden by selecting the appropriate data type from the radio buttons provided. For lists, a global delimiter must be specified (i.e., all cells in the table must use the same delimiter).

Import Networks from Web Services

From version 2.6.0, Cytoscape has a new feature called Web Service Client Manager. Users can access verious kinds of databases through this function. Please read Importing Networks and Attributes from External Database for more detail.

Edit a New Network

A new, empty network can also be created and nodes and edges manually added. To create an empty network, go to File → New → Network → Empty Network, and then manually add network components using the Editor in CytoPanel 1 (see the Editor chapter for more details).

Cytoscape can read network/pathway files written in the following formats:

• Simple interaction file (SIF or .sif format)
• Graph Markup Language (GML or .gml format)
• XGMML (extensible graph markup and modelling language).
• SBML
• BioPAX
• PSI-MI Level 1 and 2.5
• Delimited text
• Excel Workbook (.xls)

The SIF format specifies nodes and interactions only, while other formats store additional information about network layout and allow network data exchange with a variety of other network programs and data sources. Typically, SIF files are used to import interactions when building a network for the first time, since they are easy to create in a text editor or spreadsheet. Once the interactions have been loaded and network layout has been performed, the network may be saved to GML or XGMML format for interaction with other systems. All file types listed (except Excel) are text files and you can edit and view them in a regular text editor.

SIF Format

The simple interaction format is convenient for building a graph from a list of interactions. It also makes it easy to combine different interaction sets into a larger network, or add new interactions to an existing data set. The main disadvantage is that this format does not include any layout information, forcing Cytoscape to re-compute a new layout of the network each time it is loaded.

Lines in the SIF file specify a source node, a relationship type (or edge type), and one or more target nodes:

nodeA <relationship type> nodeB
nodeC <relationship type> nodeA
nodeD <relationship type> nodeE nodeF nodeB
nodeG
...
nodeY <relationship type> nodeZ

A more specific example is:

node1 typeA node2
node2 typeB node3 node4 node5
node0

The first line identifies two nodes, called node1 and node2, and a single relationship between node1 and node2 of type typeA. The second line specifies three new nodes, node3, node4, and node5; here "node2" refers to the same node as in the first line. The second line also specifies three relationships, all of type typeB and with node2 as the source, with node3, node4, and node5 as the targets. This second form is simply shorthand for specifying multiple relationships of the same type with the same source node. The third line indicates how to specify a node that has no relationships with other nodes. This form is not needed for nodes that do have relationships, since the specification of the relationship implicitly identifies the nodes as well.

Duplicate entries are ignored. Multiple edges between the same nodes must have different edge types. For example, the following specifies two edges between the same pair of nodes, one of type xx and one of type yy:

node1 xx node2
node1 xx node2
node1 yy node2

Edges connecting a node to itself (self-edges) are also allowed:

node1 xx node1

Every node and edge in Cytoscape has an identifying name, most commonly used with the node and edge data attribute structures. Node names must be unique, as identically named nodes will be treated as identical nodes. The name of each node will be the name in this file by default (unless another string is mapped to display on the node using the visual mapper). This is discussed in the section on visual styles. The name of each edge will be formed from the name of the source and target nodes plus the interaction type: for example, sourceName (edgeType) targetName.

The tag <relationship type> can be any string. Whole words or concatenated words may be used to define types of relationships, e.g. geneFusion, cogInference, pullsDown, activates, degrades, inactivates, inhibits, phosphorylates, upRegulates, etc.

Some common interaction types used in the Systems Biology community are as follows:

  pp .................. protein – protein interaction
  pd .................. protein -> DNA   
  (e.g. transcription factor binding upstream of a regulating gene.)

Some less common interaction types used are:

  pr .................. protein -> reaction
  rc .................. reaction -> compound
  cr .................. compound -> reaction
  gl .................. genetic lethal relationship
  pm .................. protein-metabolite interaction
  mp .................. metabolite-protein interaction

Delimiters

Whitespace (space or tab) is used to delimit the names in the simple interaction file format. However, in some cases spaces are desired in a node name or edge type. The standard is that, if the file contains any tab characters, then tabs are used to delimit the fields and spaces are considered part of the name. If the file contains no tabs, then any spaces are delimiters that separate names (and names cannot contain spaces).

If your network unexpectedly contains no edges and node names that look like edge names, it probably means your file contains a stray tab that's fooling the parser. On the other hand, if your network has nodes whose names are half of a full name, then you probably meant to use tabs to separate node names with spaces.

Networks in simple interactions format are often stored in files with a .sif extension, and Cytoscape recognizes this extension when browsing a directory for files of this type.

GML Format

In contrast to SIF, GML is a rich graph format language supported by many other network visualization packages. The GML file format specification is available at:

http://www.infosun.fmi.uni-passau.de/Graphlet/GML/

It is generally not necessary to modify the content of a GML file directly. Once a network is built in SIF format and then laid out, the layout is preserved by saving to and loading from GML. Visual attributes specified in a GML file will result in a new visual style named Filename.style when that GML file is loaded.

XGMML Format

XGMML is the XML evolution of GML and is based on the GML definition. In addition to network data, XGMML contains node/edge/network attributes. The XGMML file format specification is available at:

http://www.cs.rpi.edu/~puninj/XGMML/

XGMML is now preferred to GML because it offers the flexibility associated with all XML document types. If you're unsure about which to use, choose XGMML.

SBML (Systems Biology Markup Language) Format

The Systems Biology Markup Language (SBML) is an XML format to describe biochemical networks. SBML file format specification is available at:

http://sbml.org/documents/

BioPAX (Biological PAthways eXchange) Format

BioPAX is an OWL (Web Ontology Language) document designed to exchange biological pathways data. The complete set of documents for this format is available at:

http://www.biopax.org/index.html

PSI-MI Format

The PSI-MI format is a data exchange format for protein-protein interactio