Gene Set Enrichment

The Gene Set Enrichment process performs an enrichment analysis by comparing:

•	columns of continuous or binary statistics or measures, with

•	a column that defines a set of annotation categories, typically derived from an ontology hierarchy like the Gene Ontology (GO) or from lists of pathways.

This process constructs the complete set of unique categories as defined by one or two delimiters in the annotation column, and computes a Fisher exact test, PAGE Z -Score tests (Kim and Volsky, 2005), or Cochran-Armitage trend tests. The resulting p-values can be adjusted for multiplicity using a variety of methods.

What do I need?

One Input Data Set is required to successfully run the Gene Set Enrichment process. This data set must contain at least one significance variable and an annotation column. The u133a_anno.sas7bdat input data set, shown below, contains annotation information and significance data from the Affymetrix Latin Square Data , and serves as an example.

An Annotation Data Set is optional , unless the category variable is not in the input data set.

To follow this example:

•	The categories used in the enrichment are listed in the Gene_Ontology_Biological_Process column.

•	ProbF_Trt should be selected as one of the Significance Variables .

•	The annotation categories within each observation are delineated using the /// delimiter.

•	There are no secondary delimiters.

•	Enrich for genes whose expression is down-regulated, so select Smaller is more significant as the direction of significance variables.

•	Select all three enrichment tests.

•	Specify 0.01 as the significance variable cutoff for the Fisher exact test.

•	Set the number of bins for the Cochran-Armitage tests to 7 .

•	Specify a low hit threshold of 3 .

For detailed information about the files and data sets used or created by JMP Life Sciences software, see Files and Data Sets .

Output/Results

Refer to the Gene Set Enrichment output documentation for detailed descriptions of the output of this process.