calculate.correlation {MEGENA}

correlation calculation

Description

correlation analysis with FDR calculation

Usage

calculate.correlation(datExpr,doPerm = 100,doPar = FALSE,num.cores = 8,method = "pearson",
FDR.cutoff = 0.05,n.increment = 100,is.signed = FALSE,
output.permFDR = TRUE,output.corTable = TRUE,saveto = NULL)

Arguments

datExpr

gene expression data matrix

doPerm

Number of permutations to perform. If doPerm = NULL, calculates BH FDR p-values instead of permutation based FDR.

doPar

TRUE/FALSE logical variable to choose parallelization. Parallelization is utilized when BH FDR p-values are calculated for all pairs.

num.cores

number of cores to use in parallelization.

method

correlation method to be passed to cor for method argument.

FDR.cutoff

FDR threshold to output final results of significant correlations.

n.increment

When permutation is utilized, 0 <= |rho| <= 1 is broken down into n.increment to map each |rho| cutoff to respective FDR.

is.signed

TRUE/FALSE to indicate using signed/unsigned correlation.

output.permFDR, output.corTable

TRUE/FALSE to choose to output permutation indices and FDR table.

saveto

folder to output results.

Details

If doPar = TRUE, then num.cores are registered for PCP.

Value

output is three column edgelist data.frame, third column being the weight.

Author(s)

Won-Min Song

Examples

# test simplest case of planar network (a 3-clique).
data(Sample_Expression)
calculate.correlation(datExpr[1:100,],doPerm = 5)