champ.DMR: Applying Bumphunter or ProbeLasso Algorithms to detect Different Methylation Regions in a beta valued Methylation Dataset.

Description

Applying Bumphunter or ProbeLasso Algorithms to Estimate regions for which a genomic profile deviates from its baseline value. Originally implemented to detect differentially methylated genomic regions between two populations. By default, we recommend user do champ.DMR on normalized beta value on two populations, like case to control. The function will return detected DMR and estimated p value. The two algorithms specified in this function is different, while Bumphunter calcuated averaged candidate bumps methylation value between case and control, ProbeLasso need Different Methylated Probes (DMP) from champ.MVP as input parameter and find DMRs around those DMPs. Thus parameters is different for two algorithms.

Usage

champ.DMR(betaNorm=myNorm$beta,design=myLoad$pd$Sample_Group,maxGap=300, cutoff=0.5,minProbes=7,smooth=TRUE,smoothFunction=loessByCluster, useWeights=FALSE,permutations=NULL,B=250,pickCutoff=FALSE, pickCutoffQ=0.99,nullMethod="bootstrap",verbose=TRUE,cores=3,arraytype="450K", method = "Bumphunter",resultsFile=mylimma,meanLassoRadius=375,minSigProbesLasso=3,minDmrSep=1000, minDmrSize=50,adjPvalProbe=0.05,adjPvalDmr=0.05,pData=myLoad$pd)

Arguments

betaNorm

Methylation beta valueed dataset user want to do RefFreeEWAS. We recommend to use normalized beta value. In champ.DMR function, beta value will be transformed to M value. NA value is NOT allowed into this function, thus user may need to do some imputation work beforehead. This parameter is essential for both two algorithms.

arraytype

Assign types of manifest shall be used to do DMR detection, "EPIC" or "450K" are provided. Annotation of probes is playing an important role in clustering probes across genome. This parameter is both important and available for both Bumphunter and ProbeLasso functions.

method

Specify the method users want to use to do DMR detection. There are two options: "Bumphunter" or "ProbeLasso". The default value is "Bumphunter".

design

Design vector (subjects x covariates). This parameter MUST be a vecter. Though characters, factor and numeric are all allowed because inside the function, character covariates will be transformed into numeric, we still recommend user input numeric deigned covariates vector. This parameter is specified for Bumphunter algorithm.

cutoff

A numeric value. Values of the estimate of the genomic profile above the cutoff or below the negative of the cutoff will be used as candidate regions. It is possible to give two separate values (upper and lower bounds). If one value is given, the lower bound is minus the value. This parameter is specified for Bumphunter.

pickCutoff

Should bumphunter attempt to pick a cutoff using the permutation distribution? This parameter is specified for Bumphunter.

pickCutoffQ

The quantile used for picking the cutoff using the permutation distribution. This parameter is specified for Bumphunter.

maxGap

If cluster is not provided this maximum location gap will be used to define cluster via the clusterMaker function. This parameter is specified for Bumphunter.

minProbes

Threshold to filtering clusters with too few probes in it. After region detection, champ.DMR will only select probes in clusters contain more than minsample probes to continue the program. This parameter is specified for Bumphunter.

nullMethod

Method used to generate null candidate regions, must be one of ‘bootstrap’ or ‘permutation’ (defaults to ‘permutation’). However, if covariates in addition to the outcome of interest are included in the design matrix (ncol(design)>2), the ‘permutation’ approach is not recommended. See vignette and original paper for more information. This parameter is specified for Bumphunter.

smooth

A logical value. If TRUE the estimated profile will be smoothed with the smoother defined by smoothFunction, This parameter is specified for Bumphunter.

smoothFunction

A function to be used for smoothing the estimate of the genomic profile. Two functions are provided by the package: loessByCluster and runmedByCluster. This parameter is specified for Bumphunter.

useWeights

A logical value. If TRUE then the standard errors of the point-wise estimates of the profile function will be used as weights in the loess smoother loessByCluster. If the runmedByCluster smoother is used this argument is ignored. This parameter is specified for Bumphunter.

An integer denoting the number of resamples to use when computing null distributions. This defaults to 0. If permutations is supplied that defines the number of permutations/bootstraps and B is ignored. The default value is 250, This parameter is specified for Bumphunter.

permutations

is a matrix with columns providing indexes to be used to scramble the data and create a null distribution when nullMethod is set to permutations. If the bootstrap approach is used this argument is ignored. If this matrix is not supplied and B>0 then these indexes are created using the function sample. This parameter is specified for Bumphunter.

verbose

logical value. If TRUE, it writes out some messages indicating progress. If FALSE nothing should be printed. This parameter is specified for Bumphunter.

cores

The embeded DMR detection function, bumphunter, could automatically use more parallel to accelerate the program. User may assgin number of cores could be used on users's computer. The default value is 3. User may use detectCore() function to detect number of cores in total. This parameter is specified for Bumphunter.

resultsFile

Different Methylated Probes (DMP) detected from champ.MVP() function, which used limma function to find all CpGs show significant different methylation value. It's a MUST provided parameter for ProbeLasso algorithm.

meanLassoRadius

Radius around each DMP to detect DMR, the default value is 375.

minSigProbesLasso

The minimum number of significant probes to be captured in lasso, default = 3.

minDmrSep

The minimum seperation (bp) between neighbouring DMRs, default = 1000.

minDmrSize

The minimum DMR size (bp), default = 50.

adjPvalProbe

The minimum threshold of significance for probes to be includede in DMRs, default = 0.05

adjPvalDmr

This is the significance threshold for including DMRs in the final DMR list.

pData

This data.frame includes the information from the sample sheet. The default assumes you ran champ.load and saved the output to "myLoad".

Value

myDmrs: Different Methylation Regions detected by champ.DMR. For different algorithms, myDmrs would be in different structure.
myDmrProbes: Different Methylated Probes detected in DMRs.

References

Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD and Irizarry RA (2014). "Minfi: A flexible and comprehensive Bioconductor package for the analysis of Infinium DNA Methylation microarrays." Bioinformatics, 30(10), pp. 1363-1369. http://doi.org/10.1093/bioinformatics/btu049. Butcher LM and Beck S (2015). "Probe Lasso: A novel method to rope in differentially methylated regions with 450K DNA methylation data." Methods, 72, pp. 21-28. http://doi.org/10.1016%2Fj.ymeth.2014.10.036.

Examples

Run this code

	myLoad <- champ.load(directory=system.file("extdata",package="ChAMPdata"))	
    myNorm <- champ.norm(norm="NONE")
    myDMR <- champ.DMR(B=10)

Run the code above in your browser using DataLab