cea {RedeR}R Documentation

Co-expression analysis.

Description

Simple function for correlation analysis. This function computes a null distribution via permutation and returns the significant correlation values.

Usage

cea(x, sig=0.01, padj.method="fdr", cor.method="spearman", nper=1000, 
regulators=NULL, plotcea=TRUE, ...)

Arguments

x

A matrix or data frame. Varibles should be on rows, samples on columns. The 'x' matrix should be named, with unique row names.

sig

Significance threshold.

padj.method

Correction method passed to "p.adjust" function.

cor.method

Correlation method passed to "cor" function.

nper

Number of permutations.

plotcea

Logical value, option to plot density and the null distributions.

regulators

An optional character vector listing varibles (e.g. genes) regarded as regulators.

...

Additional arguments passed to plotcea option.

Details

Additional arguments:

n.breaks

If plotcea=TRUE, 'n.breaks' sets the number of histogram breaks (Default=100 <integer>).

plotnull

If plotcea=TRUE, 'plotnull' sets whether to plot the null distribution (Default=TRUE <logical>).

avnull

If plotcea=TRUE, 'avnull' takes the average null distribution (Default=TRUE <logical>).

nullcol

If plotcea=TRUE, 'nullcol' sets the color of the null distribution (Default="black" <character>).

Value

An adjacency matrix with significant correlation values.

Author(s)

Mauro Castro

See Also

cor p.adjust

Examples

#--- Load a a gene expression matrix
data(ER.deg)
x <- ER.deg$exp

#--- NOTE1: 'x' should have varibles on rows and samples on columns!
#--- NOTE2: 'x' should be named, with unique names on rows.
x[1:3,1:3]
#      GSM286031 GSM286032 GSM286033
#1      7.106765  7.225175  7.217780
#10005  8.108632  8.012673  7.932592
#10007 10.225053 10.068826 10.236548

#--- Get a subset, with 100 genes, for a quick demonstration!
x <- x[sample(1:nrow(x))[1:100], ]

## Not run: 
#--- Run correlation analysis, estimating significant 
#--- associations by permutation.
#--- NOTE: set 'nper' for at least 1000
res <- cea(x=x, nper=100)

## End(Not run)

[Package RedeR version 1.42.0 Index]