globalTest: Test whether at least one CpG is differentially methylated in...
In BiSeq: Processing and analyzing bisulfite sequencing data

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

This method is a wrapper for conveniently invoking the globaltest method gt on a BSrel-class object. The globaltest can be applied to test against a high dimensional alternative in various regression models. E.g., it can be used to test whether at least one CpG is differentially methylated between two groups.

1	globalTest(response, alternative, ...)

`response`	The response vector of the regression model. May be supplied as a vector or as a `formula` object. In the latter case, the right hand side of `response` defines the null hypothesis. The default null hypopthesis is `~1`, i.e. only an intercept.
`alternative`	An object of `BSrel-class` defining the alternative. The CpGs are used as explanatory variable in the alternative regression model. The null hypothesis is that the coefficients of all CpGs are zero. If `response` is given as `formula`, `colData(alternative)` is used to obtain the respective data.
`...`	Other arguments passed to the `gt` method. The argument `subsets` can be given as `GRanges-class` object. Then, the globaltest is applied for each region using only the CpGs lying within the respective region. This is useful for, e.g., testing all promoter regions within function call.

For details see the documentation of the gt method in package globaltest.

The function returns an object of class gt.object. Several operations and diagnostic plots for this class are provided by the globaltest package.

Hans-Ulrich Klein

Goeman, J. J., van de Geer, S. A., and van Houwelingen, J. C. (2006). Testing against a high-dimensional alternative. Journal of the Royal Statistical Society Series B- Statistical Methodology, 68(3):477-493.

link{gt}, link{BSrel}

data(rrbs)
rrbs <- rawToRel(rrbs)
regions <- GRanges(IRanges(start=c(850000, 1920000, 500), end=c(879000, 1980000, 600)),
                   seqnames=c("chr1", "chr2", "chr3"))

globalTest(group~1, rrbs)
globalTest(group~1, rrbs, subsets=regions)