addCorrelation.R: Correlation, old version
In mariavica/mircomb: An R package for analyzing miRNA-mRNA interactions

Description Usage Arguments Details Value Note See Also Examples

Correlation, old version

1
2
3

addCorrelation.R(obj, method = "pearson", subset.miRNA = obj@sig.miRNA,
 subset.mRNA = obj@sig.mRNA, common = NULL, d.influences = FALSE,
 alternative = "two.sided")

`obj`	a `corObject`.
`method`	method used for computing correlation: `"pearson"` or `"spearman"`.
`subset.miRNA`	Optional, character vector with the names of the miRNAs to correlate. It is recommended that miRNAs are added using `addSig` function.
`subset.mRNA`	Optional, character vector with the names of the mRNAs to correlate. It is recommended that mRNAs are added using `addSig` function.
`common`	Optional, character vector with the names of the samples to correlate (the samples must appear in both miRNA and mRNA datasets.)
`d.influences`	compute a matrix with the Cook's Distance of each sample in each miRNA-mRNA correlatio.
`alternative`	specification of the alternative hypothesis: `"less"` (default), `"two-sided"` or `"greater"`.

A more complete version of the addCorrelation function, but significantly slower (specially kendall correlation). Use always addCorrelation function whenever it is possible.

If d.influences = TRUE, a 3-dimension matrix is added to the info slot, labeled "influenting.sample". First dimension: miRNA names; second dimension: mRNA names; third dimension: sample names; fill: Cook's Distance for a specific sample in a specific miRNA-mRNA linear regression (defined by the dimension label-names).

A corObject with the slots "cor" and "pval" filled. Optionally, a matrix named "influenting.sample" is added to the info slot.

This function can take a long time to complete when is applied to large datasets.

corObject-class,cor, addCorrelation, addSig, cooks.distance

data(data.obj)

data.obj<-addSig(data.obj,"miRNA",manual=c("hsa-let-7c","hsa-miR-122"))
data.obj<-addSig(data.obj,"mRNA",manual=c("A1BG","A1CF"))


data.obj.correlated<-addCorrelation.R(data.obj, method="pearson", alternative="less",
 d.influences=TRUE)

data.obj.correlated@cor
data.obj.correlated@pval
data.obj.correlated@info[["influenting.sample"]]