gseaResES: Read GSEA statistic for pathway fingerprinting

In bedapub/ribiosGSEA: Gene-Set Enrichment Analysis Tools in Ribios

Read GSEA statistic for pathway fingerprinting

Description

Read GSEA statistics (log-transformed q-value [q], Enrichment Score [ES], or normalized Enrichement Score [NES]) to profile pathway activitities.

Usage

gseaResQvalue(file, threshold = 1e-04, log = FALSE, posLog = FALSE)

gseaResES(file, normalized = FALSE)

Arguments

file	GSEA output tab-delimited file, usually with the file name ‘gsea\_report\_for.*\_pos\_.*\.xls’ or ‘gsea\_report\_for.*\_neg\_.*\.xls’. Located in GSEA output directory.
threshold	Valid for q value: what is the minimum threshold of q-value (FDR)? It can be set to the number of permutation tests divided by 1. By default 1/10000
log	Valid for q value: whether the FDR q value should be transformed by base-10 (log10) logarithm. By default FALSE
posLog	Valid for q value: whether the logged FDR q value should be negated to get positive value.This is useful when the sign of q is used to distinguish between positive and negative enriched pathways. By default FALSE.
normalized	Valid for enrichment score: if set to TRUE, normalized enrichment score (nes) will be returned instead of (es). By default set to FALSE

Details

In many cases we want to extract pathway signatures from a set of experiments. Both gseaResQvalue and gseaES can read GSEA output files and extract desired statistic: q-value, ES or NES.

See the GSEA document for definitions of the three values. For comparing a few conditions to another, we recommend using q-value. For large-scale comparisons between pathways (or other gene signatures), we have found ES very useful. It is adviced to choose proper statistic to extract pathway signatures only when you are sure of the aim. Using any statistic without good reasoning may as always lead to wrong intepretations of the data.

These functions are usually not directly called by end-users. See gseaFingerprint and link{gseaFingerprintMatrix} instead.

Value

A data.frame with two columns: name and value. The column name contains gene signatures (e.g. pathways), and value contains the statistic.

Functions

• gseaResQvalue(): The function to extract the Q-value

  Extract Q-values from GSEA result file

Author(s)

Jitao David Zhang <[e-mail]>, with input from Martin Ebeling, Laura Badi and Isabelle Wells.

References

GSEA documentation http://www.broadinstitute.org/gsea/doc/GSEAUserGuideFrame.html

See Also

End-users will probably find gseaFingerprint and link{gseaFingerprintMatrix} more useful, since they operate on the level of GSEA result directories, instead of single output tab-delimited files.

Examples

gseaDirZip <- system.file(package="ribiosGSEA","extdata/gseaDirs.zip")
tmpDir <- tempdir()
utils::unzip(gseaDirZip, exdir=tmpDir)
gseaDir <- file.path(tmpDir, "gseaDirs")
gseaFile <- file.path(gseaDir,
   "VitaminA_24h_High",
   "gsea_report_for_na_neg_1336489010730.xls")

gseaQ <- gseaResQvalue(gseaFile)
gseaLogQ <- gseaResQvalue(gseaFile, log=TRUE)
gseaQscore <- gseaResQvalue(gseaFile, log=TRUE, posLog=TRUE)

gseaEs <- gseaResES(gseaFile)
gseaNes <- gseaResES(gseaFile, normalized=TRUE)

bedapub/ribiosGSEA documentation built on March 30, 2023, 3:26 p.m.