reformData: Reform M-values into a two-column matrix.

In DMRMark: DMR Detection by Non-Homogeneous Hidden Markov Model from Methylation Array Data

Description

Reform M-values into a matrix with two columns representing matched control and case groups. It concatenate M-values pair-by-pair based on the design matrix.

Usage

reformData(mv, pd=NULL)

Arguments

mv	The input M-values matrix, NA is not allowed.
pd	A design matrix, which can be generated by 'stats::model.matrix'. If the M-values are totally paired or single paired, just leave it to be NULL.

Value

A matrix with two columns representing matched control and case groups. If a sample has no paired sample in another group (say group B), then the values in group B will be represented by NA.

Author(s)

Linghao SHEN <[e-mail]>

Examples

	# Assume the values come from Tumor is 10 larger than those from Normal.
	
	# The case with totally paired data 
	mv1 <- matrix(1:20,5)
	reformData(mv1)
	
	# The case with One more sample from Tumour group
	# The second Tumour sample is the extra one
	mv2 <- matrix(1:25,5)
	mv2[,2] <- mv2[,2] + 5 
	patient <- factor(c(1,3,1:3))
	type = c(rep("Normal",2),rep("Tumour",3))
	pd <- model.matrix(~patient + type + 0)
	reformData(mv2, pd)

DMRMark documentation built on May 2, 2019, 1:53 p.m.