Gsea at the Broad

From GeneSetEnrichmentAnalysisWiki

GSEA Home | Downloads | Molecular Signatures Database | Documentation | Contact

This page describes how to use GSEA from the command line on the Broad UNIX system.
It applies only to researchers at the Broad institute. 

GSEA Software

For GSEA version 2:

• The latest jar file is available at:

  /xchip/projects/xtools/gsea2.jar

• The following utility makes usage easier (it sets up CLASSPATH etc):

  /xchip/projects/xtools/run

• GSEA requires Java1.5, which is available at:

  /usr/java/java1.5/bin/java

Command Lines

If your dataset and gene sets already match, use the following command line to run GSEA:

/xchip/projects/xtools/run gsea -res foo.gct -cls foo.cls -gmx foo.gmx -collapse false

java -Xmx1024m -cp /xchip/projects/xtools/gsea2.jar xtools.gsea.Gsea -res foo.gct -cls foo.cls -gmx foo.gmx -collapse false             <span style="font-family: arial,verdana,sans-serif;"><br /></span>

If you want to use gene symbol based gene sets and your dataset is in some probe set format:

/xchip/projects/xtools/run gsea -res foo.gct -cls foo.cls -gmx foo.gmx -chip HG_U133A

java -Xmx1024m -cp /xchip/projects/xtools/gsea2.jar xtools.gsea.Gsea -res foo.gct -cls foo.cls -gmx foo.gmx -chip HG_U133A

Note: For the -chip parameter, specify just the name and not the path. For the full list of supported chips, see /xchip/projects/vdb/annotations.

Analysis Results

• By default, the output is written to a folder in the current working directory with a date prefix.
• Change the output folder by specifying the -out parameter.
• Change the name of the report by specifying the -rpt_label parameter.