calculate.GSEA: Calculate 2-sided statistics based on the GSEA algorithm

Description

Calculates the 2-sided statistics based on the GSEA algorithm.

Usage

calculate.GSEA(tab, phenotype, gsList, nsim = 1000, verbose = FALSE, alwaysUseRandomPerm = FALSE)

Arguments

tab

a numeric matrix of expression values, with the rows and columns representing probe sets and sample arrays, respectively

phenotype

a numeric or character vector indicating the phenotype

gsList

a list containing three vectors from the output of the selectGeneSets function

nsim

an integer indicating the number of permutations to use

verbose

a boolean to indicate whether to print debugging messages to the R console

alwaysUseRandomPerm

a boolean to indicate whether the algorithm can use complete permutations for cases where nsim is greater than the total number of unique permutations possible with the phenotype vector

Value

ngs: number of gene sets
nsim: number of permutations performed
t.set: a numeric vector of Tk statistics
t.set.new: a numeric vector of NTk statistics
p.null: the proportion of nulls
p.value: a numeric vector of p-values
q.value: a numeric vector of q-values

Details

This function assumes 2 distinct types of phenotypes in the data. It calculates a variant of the GSEA statistics (Mootha et al.) with the following modifications: (a) GSEA was changed from a 1-sided to a 2-sided approach. (b) The 2-group t-statistics is used as the difference metric.

The function also normalizes the GSEA statistic and calculates the corresponding q-values for each gene set as described in Tian et al. (2005) The function's output can be used for further analysis in other functions such as rankPathways.NGSk or getPathwayStatistics.NGSk.

References

Mootha V.K., Lindgren C.M., Eriksson K.F., Subramanian A., Sihag S., Lehar J., Puigserver P., Carlsson E., Ridderstrale M., Laurila E., Houstis N., Daily M.J., Patterson N., Mesirov J.P., Golud T.R., Tamayo P., Spiegelman B., Lander E.S., Hirshhorn J.N., Altshuler D., Groop L.C. (2003) PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nature Genetics, 34, 267-73. Tian L., Greenberg S.A., Kong S.W., Altschuler J., Kohane I.S., Park P.J. (2005) Discovering statistically significant pathways in expression profiling studies. Proceedings of the National Academy of Sciences of the USA, 102, 13544-9.

http://www.pnas.org/cgi/doi/10.1073/pnas.0506577102