randomDiana2means: Generate null-distributions of DLD-scores
In adSplit: Annotation-Driven Clustering

Description Usage Arguments Details Value Author(s) See Also Examples

View source: R/adSplit.R

Draws a number of random sets of probe-sets consisting of the needed size and applies diana2means to compute DLD scores.

1 2	randomDiana2means(nprobes, data, chip, ndraws = 10000, ngenes = 50, ignore.genes = 5)

`nprobes`	the size of gene sets.
`data`	a matrix of expression data, rows correspond to genes, columns to samples.
`chip`	the name of the used chip.
`ndraws`	the number of DLD scores computed.
`ngenes`	the number of genes used to compute DLD scores (passed to `diana2means`).
`ignore.genes`	the number of best scoring genes to be ignored when computing DLD scores (passed to `diana2means`)

This function uses drawRandomPS to draw ndraws gene sets. On these it applies diana2means to determine a null-distribution of DLD-scores.

A vector of DLD-scores.

Joern Toedling, Claudio Lottaz

drawRandomPS, diana2means

# prepare data
library(vsn) 
library(golubEsets) 
data(Golub_Merge) 

# generate DLD scores
scores <- randomDiana2means(20, exprs(Golub_Merge), "hu6800", ndraws = 500)