compute.pairw.cor.meta: Function to compute pairwise correlations in a...

In genefu: Computation of Gene Expression-Based Signatures in Breast Cancer

Description

This function computes meta-estimate of pairwise correlation coefficients for a set of genes from a list of gene expression datasets.

Usage

compute.pairw.cor.meta(datas, method = c("pearson", "spearman"))

Arguments

datas	List of datasets. Each dataset is a matrix of gene expressions with samples in rows and probes in columns, dimnames being properly defined. All the datasets must have the same probes.
method	Estimator for correlation coefficient, can be either pearson or spearman.

Value

cor	Matrix of meta-estimate of correlation coefficients with probes in rows and prototypes in columns.
cor.n	Number of samples used to compute meta-estimate of correlation coefficients.

Author(s)

Benjamin Haibe-Kains

See Also

map.datasets, compute.proto.cor.meta

Examples

## load VDX dataset
data(vdxs)
## load NKI dataset
data(nkis)
## reduce datasets
ginter <- intersect(annot.vdxs[ ,"EntrezGene.ID"], annot.nkis[ ,"EntrezGene.ID"])
ginter <- ginter[!is.na(ginter)][1:30]
myx <- unique(c(match(ginter, annot.vdxs[ ,"EntrezGene.ID"]),
  sample(x=1:nrow(annot.vdxs), size=20)))
data2.vdxs <- data.vdxs[ ,myx]
annot2.vdxs <- annot.vdxs[myx, ]
myx <- unique(c(match(ginter, annot.nkis[ ,"EntrezGene.ID"]),
  sample(x=1:nrow(annot.nkis), size=20)))
data2.nkis <- data.nkis[ ,myx]
annot2.nkis <- annot.nkis[myx, ]
## mapping of datasets
datas <- list("VDX"=data2.vdxs,"NKI"=data2.nkis)
annots <- list("VDX"=annot2.vdxs, "NKI"=annot2.nkis)
datas.mapped <- map.datasets(datas=datas, annots=annots, do.mapping=TRUE)
## compute meta-estimate of pairwise correlation coefficients
pairwcor <- compute.pairw.cor.meta(datas=datas.mapped$datas, method="pearson")
str(pairwcor)

genefu documentation built on Jan. 28, 2021, 2:01 a.m.