Gene Enrichment Profiler

Files for download

	Primary cells U133A	Primary+Cancer cells U133A
Mean expression data matrix	Download(gzip)	Download(gzip)
Standard deviation expression data matrix	Download(gzip)	Download(gzip)
PMA data matrix (Average of PMA score where Present=1 and Absernt/Marginal=0)	Download(gzip)	Download(gzip)
Enrichment data matrix	Download(gzip)	Download(gzip)
Probe annotation (by Affymetrix)	Download probe to HUGO symbols (gzip)
Probe annotation (our own mapping)	Download probe to HUGO symbols (gzip)

Source code

The R source code is available here. To use this code in R you must install bioconductor.
To generate enrichment scores in R you need a body atlas size data set with replicates in each group. Normalize and filter the data set as you see fit. The group membership of each replicate is determined by a character vector we refer to as groups in the source code. Replicates must have the same name and names must be valid names in R (i.e., no spaces, slashes, etc.). A group of 6 samples with duplicates could look like this:
groups<-c("EScells", "EScells", "Tcells", "Tcells", "Bcells", "Bcells").
The order of the groups must be identical to the experssion data matrix. The algorithm will assume the first 2 items are duplicate samples of ES cells, etc. To run the code simply type:
enrichment<-GenerateEnrichmentScore(expressionMatrix, groupsVector)

The coefficient and p-value matrices are computed on the fly and not saved. Here is an example of the coefficient matrix and p-values generated for ES cells in the primary cell data set.