The latest versions of the GenePattern components are listed in the Release Notes.

GenePattern platform compatibility information is provided in the Release Notes.

GenePattern hardware requirements are provided in the Release Notes.

GenePattern browser support information is provided in the Release Notes.

In addition to this FAQ, the GenePattern team provides the following online resources:

• The Concepts Guide provides a brief introduction to GenePattern: its primary objects (modules, pipelines, suites) and its client-server architecture.
• The Tutorial provides a hands-on tour of the Web Client and Desktop Client.
• The Desktop Client Guide describes the Desktop Client and how to use it.
• The Web Client Guide describes the Web Client and how to use it.
• The Programmers Guide provides guidelines for writing modules and instructions for accessing GenePattern from the Java, MATLAB, and R programming environments.
• The Modules page lists the modules and pipelines in the Broad repository, with links to their documentation.
• The File Formats Guide describes all file formats and provides instructions for creating input files.
• The Release Notes describe new features in the current release.

To provide feedback or ask a question not addressed by the online resources, send email to gp-help@broad.mit.edu.

To cite GenePattern, please use the following citation:
Reich M, Liefeld T, Gould J, Lerner J, Tamayo P, Mesirov JP (2006) GenePattern 2.0 Nature Genetics 38 no. 5 (2006): pp500-501 doi:10.1038/ng0506-500.

To cite a GenePattern analysis or visualization module, cite the GenePattern software and the original paper or other source for the module as specified in the module documentation. Documentation for each module is available on the Modules page and in the GenePattern Desktop and Web clients (click Help when prompted to enter the module's parameters).

If you are a member of the press and need additional information about GenePattern, please contact our Public Relations department by calling 617.452.3000.

To install GenePattern, go to GenePattern Download and follow the instructions for your operating system.

To uninstall GenePattern, use the utility provided as part of the GenePattern installation. If the GenePattern uninstall utility is unavailable, deleting the GenePattern installation folder removes all GenePattern files other than the desktop icons.

Mac users: If R2.5 is not already installed, GenePattern installs it in the /Library/Frameworks/R.framework/Versions/2.5 folder. Uninstalling GenePattern does not uninstall R. To uninstall R, use the utility provided by R.

Simply install the new version of GenePattern into the same directory as your previous version. Do not uninstall first. It is unnecessary and will delete your existing modules, pipelines and suites. When you overwrite the previous version:

• Existing modules, pipelines and suites are preserved.


• The following settings are read from your existing genepattern.properties file and displayed as default values for your new installation: settings for R, Java, Perl, LSID Authority, proxy settings, HSQL database URL and port, and file purge frequency and time.


• The default value for the webserver (Tomcat) port used by the GenePattern server is always 8080. If your existing installation uses a different port number, you can specify that port number during the installation.


• The require.password setting from your existing genepattern.properties file is preserved in your new installation.


• Backup copies of the following configuration files are created: genepattern.properties[.save], permissionMap.xml[.saved], and userGroups.xml[.saved]. To recreate your previous settings after installing GenePattern, compare the saved files with the newly installed files and modify the new files as necessary. Do not replace the newly installed configuration files with the saved copies.

Installing GenePattern 3.1 installs R2.5 and sets the full path to R2.5. Installing GenePattern 3.1 into the same directory as GenePattern 3.0, preserves your existing R2.0.1 installation and full path to R2.0.1. This ensures that you have access to GenePattern modules written for either R2.0.1 or R2.5. For more information, see GenePattern 3.1 Release Notes.

Click here for the latest release notes.

Yes. Source code for GenePattern and its modules is available under the GenePattern software license. For access to the source code, email the GenePattern team.

No. The R, Perl, and Java installations that come with the GenePattern are installed within the GenePattern directory and do not affect any other versions that you may currently have.

You can configure GenePattern to work with versions of R/Perl/Java already on your machine; however, the versions of R, Perl, and Java that are bundled with GenePattern are the ones that have been fully tested. We cannot guarantee that other versions will work.

If you install a GenePattern server without the Java VM, choosing instead to use a Java VM that you have already installed, you must ensure that the file tools.jar (provided by SUN seperately from the JRE and JDK) is on your classpath before you can run the MatlabComponentRuntime (MCR) Installer. When you install a GenePattern server with an included VM, the GenePattern installation does this for you.

GenePattern supports the Basic Latin character set. Characters other than those in the Basic Latin character set may not be displayed correctly. Asian character sets are not currently supported.

The most likely reason is that when the installer finished, you missed the step to install the modules that came up in a browser window after the installer closed. To install the modules:

1. Start the GenePattern Web Client (http://localhost:8080/gp is usually the URL).
2. Click Modules & Pipelines>Install from repository. GenePattern displays all of the new and/or updated modules for your system checked.
3. Click Install Checked to install them onto your GenePattern server.

If you did not indicate that you were behind a web proxy/firewall when you installed GenePattern, you must update the proxy settings for your server before you can install the modules:

1. Start the GenePattern Web Client (http://localhost:8080/gp is usually the URL).
2. Click Administration>Server Settings to display the server settings.
3. Click Proxy to display the proxy settings.
4. On the Proxy Settings page, enter the hostname and port of your web proxy server. If you do not know them, contact your IT help desk to get the values. If you need to log into the proxy server, also enter your username and password (these will NOT be saved to a file and will need to be reentered following a server restart next time you want to connect).
5. Click Save to update the proxy settings.
6. Click Modules & Pipelines>Install from Repository to install the modules.

If you still cannot connect to the repository, email us at gp-help@broad.mit.edu.

You need to use the war file installation. Instructions are available here.

If you already have a server such as Tomcat running on this port, you need to install the GenePattern server on a different port to avoid conflicts. When you install the GenePattern Desktop Client, you must specify the the port on which the GenePattern server is installed; by default, the installer selects port 8080. If you install the GenePattern server on a different port, when you install the GenePattern Desktop Client, in the "GenePattern server configuration" window choose that port as the "webserver port" parameter.

Mac OS X

• Control-click on the file GenePattern/Tomcat/StartGenePatternServer (server) or the GenePatterClient/GenePattern Client icon (client).
• Select 'Show Package Contents' from the pop-up menu. The Contents directory should open in the finder.
• In the Contents directory, double-click the Info.plist file. This should open the Property List Editor program.
• Add the child 'VMOptions' under the 'Java' node.
• Change the Class of the added VMOptions node to 'Array'
• Add the child with Class 'String' with the value -Xmx512M. You can replace the value 512 with the maximum amount of memory in MB that you want the GenePattern Client to use.

• Edit the file in the GenePattern directory called StartGenePatternServer.lax (server) or in the GenePatternClient directory called GenePattern Client.lax (client).
• In either of these files, look for the entries noted below and increase these values by doubling up to the max memory size of the machine you are using. (Note: Windows limits the total space available to a process to 2 GB. Some of that is used for overhead, so slightly less is really available to the JRE.)
  • lax.nl.java.option.java.heap.size.initial
  • lax.nl.java.option.java.heap.size.max

You can use the Web Client to change the amount of memory allocated to modules written in Java and R. For modules written in other languages, you may or may not be able to change the amount of memory that is allocated to the module.

To change the memory allocated to modules:

1. Start the GenePattern Web Client (http://localhost:8080/gp is usually the URL).
2. Click Administration>Server Settings to display the server settings.
3. Click Programming Languages. The Programming Language Options section of this page specifies the options (flags) passed on the module command line.
4. In the Java VMOptions field, modify the Java -X flag to increase the memory allocated for Java modules run on the server (non-visualizer modules); for example, -Xmx1024M doubles the amount of memory allocated. You can specify up to the maximum memory size of your server machine.
5. In the Java Visualizer VMOptions field, modify the Java -X flag to increase the memory allocated by default for modules run on the client (visualizer modules).
   GenePattern users can customize the memory allocated to visualizer modules based on the amount of memory available on their desktop PCs. To customize the memory allocated to visualizer modules run on your own desktop: select My Settings, select Visualizer Memory, and modify the Java -X flag to specify the amount of memory to be allocated to visualizer modules run on your desktop.
6. In the R Options field, add the --max-mem-size flag to specify the maximum amount of memory to be used by R modules running on the server; for example, --max-mem-size=1G allocates a gigabyte. You can specify up to the maximum memory size of your server machine. The --max-mem-size flag affects only Windows operating systems.
7. Click Save to save your changes.

Changes to Java VMOptions and R Options take effect when you stop and restart the server. Changes to the Java Visualizer VMOptions take effect the next time you run a visualizer module.

Yes. If you are running more than one installation of GenePattern on the same machine, you must make sure that the port numbers for the GenePattern server and the HSQL server are unique to each installation. The Tomcat server listens on two ports, 8080 (requests) and 8005 (shutdown) by default, and the HSQL server listens on port 9001. All 3 ports need to be modified on the second copy of Tomcat. For example, you can set the GenePattern server port to 8080 and 8005 on one install and 8081 and 8086 on the other, and set the HSQL port to 9001 on one and 9002 on the other. You can configure these port numbers when you are installing the server.

Using a Queuing System in the GenePattern Web Client Guide discusses how to integrate GenePattern with queuing systems such as the Load Sharing Facility (LSF) and the Sun Grid Engine (SGE).

Choose one hostname for the GenePattern server; for example, http://servername.domainname.edu:8080/gp/. Edit the genepattern.properties file and set the following properties:

• GenePatternURL=http\://servername.domainname.edu:8080/gp/
• GENEPATTERN_PORT=8080
• gpServerHostAddress=servername.domainname.edu
• fqHostName=servername.domainname.edu
• fullyQualifiedHostName=servername.domainname.edu

Session timeout is set in the Tomcat configuration file of the GenePattern server. To modify this setting for a local GenePattern server:

1. Edit the Tomcat configuration file in the GenePattern installation directory: GenePatternServer/Tomcat/conf/web.xml.
2. Modify the session-timeout property. Enter a value in minutes, where 0 disables session timeout. For example, to set the timeout to one day:

   <session-config>
      <session-timeout>1440</session-timeout>
   </session-config>

3. Save the Tomcat configuration file.
4. Restart the GenePattern server.

On the public GenePattern server, session timeout is set to four hours and cannot be modified by a user.

Yes. Most GenePattern analyses can run on 2-channel or ratio-based data as easily as on single channel or absolute value data. To run 2-channel data in GenePattern, do the following:

• Convert your ratio-based data to a GenePattern GCT file. This tab-delimited text file format contains features (genes or probes), samples, and a computed ratio value for each feature in each sample.


• GenePattern modules cannot analyze files with missing values. If your data has missing values, one way to address the issue is to use the ImputeMissingValues.KNN module to impute the missing values.

Your data is now in a GCT file that can be analyzed by most GenePattern modules. (If you want to use non-negative matrix factorization (NMF) and your data contains negative values, see the NMF note in the Modules & Pipelines section below.)

Ratio values for cDNA data can be computed using a variety of methods. How the ratios are computed determines whether it is possible to create a class (CLS) file for the cDNA ratio data. For example:

• If ratios for all samples are computed against a common reference, as shown below, each sample can be assigned a distinct class and it is possible to create a class (CLS) file.

  normal sample (Cy3) / common reference (Cy5) = phenotype 1
  treated sample (Cy3) / common reference (Cy5) = phenotype 2



• If ratios are computed by comparing conditions, as shown below, it may not be possible to create a CLS file.

  normal sample (Cy3) / treated sample (Cy5) = phenotype

If you cannot create a CLS file, you can analyze your data using modules that do not require class files (such as ConsensusClustering), but will not be able to use modules that require the CLS file (such as ComparativeMarkerSelection).

Information on file formats supported by the modules currently in GenePattern is available in File Formats.

Run your file through PreprocessDataset. Select the desired output format for your file. If you only want to convert the file type without filtering, select "no filter" as the choice for the "filter flag" parameter.

File Formats describes the file formats used in GenePattern and, where applicable, suggests methods for converting files to these formats.

The ExpressionFileCreator module converts a set of individual CEL files into an expression data set that is usable by GenePattern modules. The MAGEMLImportViewer module imports data in MAGE-ML format into GenePattern.

Right-click the project and select Remove from the context menu.

On servers running on Windows, the following input filenames will not work, regardless of the extension they are given: con, prn, aux, clock$, nul, com, com1, com2, com3, com4, lpt1, lpt2, lpt3. If your filename consists of one of these with any extension, you will not be able to save it.

Yes. Simply install the new version of GenePattern into the same directory as your previous version. See FAQ 2.2 for more information.

This generally occurs for one of two reasons:

• If the same zip file is installed twice, by two users, the second one overwrites the first one. While the bits are the same (including LSID), the ownership and privacy are subject to change and may end up hiding it from the module's original installer if the second installer installs it as private.
• The same suite cannot be installed as a "private" suite for more than one user. If you install a private suite and do not see it, it may already be installed as a private suite by another user.

Pipeline input files with spaces in their names may give file-not-found errors. If this happens, use DOS' "dir /x" command to get the 8.3 version of the directory and filename and use that instead of the long filename. If you are using a Unix-based platform, you may need to quote the filename parameters on the command line definition.

These pipelines are the actual analyses performed in the Nature paper, "MicroRNA Expression Profiles Classify Human Cancers" by Lu, Getz, Miska, et. al. Running these pipelines will exactly reproduce their results. You can also tweak the parameters used to see for yourself how their results change if they had performed any step in their analyses differently. Additional information about the paper is available at http://www.broad.mit.edu/cancer/pub/miGCM.

To run NMF on data that contains negative values, you must do the following (using the method of Kim, P. M. & Tidor, B. (2003) Genome Res. 13, 1706-1718):

• Create one dataset with all negative numbers zeroed
• Create another dataset with all positive numbers zeroed and the signs of all negative numbers removed
• Merge the two (eg. by concatenation), resulting in a dataset twice as large as the original, but with positive values only and zeros, hence appropriate for NMF.

To do this in MATLAB, you can execute the following:
anew=[max(a,0);-min(a,0)];
where a is the original data.
We are currently developing a GenePattern module to perform this operation as well.

No, you can use the two files that are created and leave the remaining input box blank. HierarchicalClustering creates a cdt file and one or two additional files: an atr file if you clustered by samples (columns), a gtr file if you clustered by genes (rows), or both atr and gtr files if you clustered by both samples and genes (columns and rows). The JavaTreeView module accepts the two or three files created by HierarchicalClustering.

The HeatMapViewer module currently does not include gene annotations with the saved image. Use the HeatMapImage module to include gene annotations.

When computing the t-test or signal to noise ratio, ClassNeighbors thresholds the standard deviation to ensure that it is at least twenty percent of the mean. Additionally, if the standard deviation is zero, ClassNeighbors sets it to 0.1.

Yes. You can use the GSEA module with the c3 (motif) gene sets. The GSEA module is documented on the Modules page.

You must install Java 3D (https://java3d.dev.java.net/binary-builds.html).

The GenePattern server itself does not connect to any database, but modules can and have been written to connect to databases and retrieve data from them including caArray (caArrayImportViewer) and Gene Expression Omnibus (GEOImporter). To connect to any database of your choice, write a simple command-line program to connect to the database and retrieve data into a file format and install this program as a module into GenePattern (see Creating Modules in the GenePattern Web Client Guide).

When creating a matlab visualizer using matlab 7.0 compiled m-code (any release before 7.4), any figures that you create in MATLAB must have the value visible set to on or they will not be drawn to the screen.

Yes. When you define the command line to run the module, you use the substitution variable <R> to reference the R configuration property, which provides the full path to the R installation used by GenePattern. To use a different R installation, create a custom property, set that property to the desired R installation, and reference that property in your module command line (see Defining the Module Command Line).

To create a custom property, in the GenePattern Web Client select Administration>Server Settings and edit the Custom settings (see Modifying Server Settings). If you publish your module, include the name and definition of the custom property in your module documentation. Other GenePattern users will need to define that custom property to use the module.

The reference guide for accessing GenePattern modules from Java, MATLAB, and R is the GenePattern Programmers Guide.

GenePattern is based on a web services API already so you may not need to create a new web service for this purpose. The WSDL for the GenePattern server is available at

To get an easy start on creating a web services client to GenePattern:

1. Start the GenePattern Web Client (http://localhost:8080/gp is usually the URL).
2. Create a one-step pipeline (see Creating Pipelines in the GenePattern Web Client Guide).
3. Select the pipeline that you just created. GenePattern displays the parameters (if any) for the pipeline.
4. At the bottom of that page, select your favorite programming language and click View Code. GenePattern generates the code required to run the pipeline.
5. Select Downloads>Programming Libraries to download the GenePattern programming library for your favorite programming language.
6. Compile and run the generated code.

You can then modify the pipeline code to do what your application needs.

For more information about the programming libraries, see the GenePattern Programmers Guide.

A pipeline or module with a period in its name cannot be called from MATLAB. When you create a pipeline using the GenePattern 3.0 Web Client, the Web Client appends .pipeline to the name. Therefore, pipelines created in the Web Client cannot be called from MATLAB.