R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Computes the components of the M3D test-statistic over one...

M3D_Single

R Documentation

Computes the components of the M3D test-statistic over one region for 2 samples

Description

Returns the two components of the M3D test-statistic - the MMD (Gretton et al. 2006) for the full data and the coverge only data, respectively. This is not intended to be called directly by the user.

Usage

M3D_Single(testData, locMx, locInds, method = "MinusCovMMD")

Arguments

`testData`	Contains the methylation data over two samples for a given region.
`locMx`	The matrix of distances between the CpG sites.
`locInds`	The indices of the non-zero entries of locMx.
`method`	This specifies whether to return the full MMD and the methylation blind MMD (focusing only on the coverage) or just the former. if method ='MinusCovMMD' it is both, all other values return just the full MMD.

Value

This returns the value of the MMD for the region between the two samples as a numeric. If method is set to 'MinusCovMMD', a list is returned of the full MMD and the coverage only MMD. Subtracting them gives the M3D test-statistic.

Author(s)

Tom Mayo t.mayo@ed.ac.uk

References

Gretton, A., Borgwardt, K. M., Rasch, M., Scholkopf, B., Smola, A. J. (2006). A kernel method for the two-sample-problem. In Advances in neural information processing systems (pp. 513-520).