bga.jackknife: Jackknife between group analysis
In SamGG/made4: Multivariate analysis of microarray data using ADE4

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

Performs one-leave-out jackknife analysis of a between group analysis as described by Culhane et al., 20002

1	bga.jackknife(data, classvec, ...)

`data`	Input dataset. A `matrix`, `data.frame` If the input is gene expression data in a `matrix` or `data.frame`. The columns contain the cases (array samples) which will be jackknifed.
`classvec`	A factor or vector which describes the classes in the training dataset
`...`	further arguments passed to or from other methods

Performs a one-leave-out cross validation of between group analysis bga. Input is a training dataset. This can take 5-10 minutes to compute on standard data gene expression matrix.

In jackknife one leave out analysis, one case (column) is removed. The remaining dataset is subjected to bga. Then the class of the case that was removed is predicted using suppl. This analysis is repeated until all samples have been removed and predicted.

A list containing

`results`	The projected co-ordinates of each sample
`summary`	A summary of number and percentage of correctly assigned samples

Aedin Culhane

Culhane et al., 2002 Between-group analysis of microarray data. Bioinformatics. 18(12):1600-8.

See Also bga, bga.suppl, suppl,bga, bca, plot.bga

data(khan)
# NOTE using a very reduced dataset (first 5 genes) to speed up results
# hence expect poor prediction accuracy
dim(khan$train)
print("using only small subset of data")
if (require(ade4, quiet = TRUE)) {
bga.jackknife(khan$train[1:5,], khan$train.classes) }