preprocessENmix: The ENmix background correction

Description Usage Arguments Details Value Author(s) References Examples

View source: R/preprocessENmix.R

Description

The ENmix background correction for HumanMethylation 450 and MethylationEPIC BeadChip. ENmix models methylation signal intensities with a flexible exponential-normal mixture distribution, and models background noise with a truncated normal distribution. ENmix will split BeadChip intensity data into 6 parts and separately model methylated and unmethylated intensities, 2 different color channels and 2 different probe types.

Usage

1
2
preprocessENmix(rgSet, bgParaEst = "oob", dyeCorr="RELIC", QCinfo=NULL, exQCsample=TRUE,
                    exQCcpg=TRUE, exSample=NULL, exCpG=NULL, nCores = 2)

Arguments

rgSet

An object of class rgDataSet, methDataSet, RGChannelSetExtended, RGChannelSet or MethylSet.

bgParaEst

Method to estimate background normal distribution parameters. Options are: "oob","est", or "neg".

dyeCorr

Dye bias correction method, "mean": correction based on averaged red/green ratio, or "RELIC": correction with RELIC method (default method), or "none": no dye bias correction.

QCinfo

If QCinfo object from function QCinfo() is provided, low quality samples (if exQCsample=TRUE) and CpGs (if exQCcpg=TRUE) will be excluded before background correction.

exQCsample

If TRUE, low quality samples listed in QCinfo will be excluded.

exQCcpg

If TRUE, low quality CpGs listed in QCinfo will be excluded.

exSample

User specified samples to be excluded before background correction

exCpG

User specified probes to be excluded before background correction

nCores

Number of cores will be used for computation

Details

By default, ENmix will use out-of-band Infinium I intensities ("oob") to estimate normal distribution parameters for modeling background noise. Option "est" will use combined methylated and unmethylated intensities to estimate background distribution parameters separately for each color channel and each probe type. Option "neg" will use 600 chip internal controls probes to estimate background distribution parameters. If rgSet if a MethylSet, then only option "est" can be selected.

Value

An object of class same with input data

Author(s)

Zongli Xu and Liang Niu

References

Zongli Xu, Liang Niu, Leping Li and Jack A. Taylor, ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip. Nucleic Acids Research 2015.

Zongli Xu, Sabine A. S. Langie, Patrick De Boever, Jack A. Taylor1 and Liang Niu, RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip, BMC Genomics. 2017

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
if (require(minfiData)) {
#rgDataSet as input
path <- file.path(find.package("minfiData"),"extdata")
rgSet <- readidat(path = path,recursive = TRUE)
#quality control information
qc<-QCinfo(rgSet)
#further excluding samples which are not in the qc$badsample list
ex_id=c("5723646053_R04C02")
#further excluding cpgs which are not in the qc$badCpG list
ex_cg=c("cg00000622", "cg00001245", "cg00001261")
mdat=preprocessENmix(rgSet,QCinfo=qc,exSample=ex_id,exCpG=ex_cg,nCores=6)


#RGChannelSet as input
mdat=preprocessENmix(RGsetEx,nCores=6)
}

ENmix documentation built on April 2, 2021, 6 p.m.