%\VignetteIndexEntry{ENmix User's Guide} %\VignetteKeywords{DNA methylation, background correction, preprocessing} %\VignettePackage{ENmix} \documentclass[12pt]{article} <>= options(width=70) @ \usepackage{Sweave} \usepackage[utf8]{inputenc} \usepackage{times} \usepackage{color, hyperref} \usepackage{fullpage} \usepackage{parskip} \usepackage{multirow} \usepackage{booktabs} \newcommand{\Rpackage}[1]{\textsf{#1}} \newcommand{\Rfunction}[1]{{\texttt{#1}}} \newcommand{\Robject}[1]{{\texttt{#1}}} \title{The \Rpackage{ENmix} User's Guide} \author{Zongli Xu, Liang Niu and Jack A Taylor} \date{Modified: October 9 2020. Compiled: \today} \begin{document} \maketitle \tableofcontents \newpage \section{Introduction} The \Rpackage{ENmix} package provides a set of quality control, preprocessing/correction and data analysis tools for Illumina Methylation Beadchips. It includes functions to read in raw idat data, background correction, dye bias correction, probe-type bias adjustment, along with a number of additional tools. These functions can be used to remove unwanted experimental noise and thus to improve accuracy and reproducibility of methylation measures. \Rpackage{ENmix} functions are flexible and transparent. Users have option to choose a single pipeline command to finish all data pre-processing steps (including quality control, background correction, dye-bias adjustment, between-array normalization and probe-type bias correction) or to use individual functions sequentially to perform data pre-processing in a more customized manner. In addition the \Rpackage{ENmix} package has selectable complementary functions for efficient data visualization (such as QC plots, data distribution plot, manhattan plot and Q-Q plot), quality control (identifing and filtering low quality data points, samples, probes, and outliers, along with imputation of missing values), identification of probes with multimodal distributions due to SNPs or other factors, exploration of data variance structure using principal component regression analysis plot, preparation of experimental factors related surrogate control variables to be adjusted in downstream statistical analysis, an efficient algorithm oxBS-MLE to estimate 5-methylcytosine and 5-hydroxymethylcytosine level; estimation of celltype proporitons; methlation age calculation and differentially methylated region (DMR) analysis. Most \Rpackage{ENmix} package can also support the data structure used by several other related R packages, such as \Rpackage{minfi}, \Rpackage{wateRmelon} and \Rpackage{ChAMP}, providing straightforward integration of ENmix-corrected datasets for subsequent data analysis. The software is designed to support large scale data analysis, and provides multi-processor parallel computing options for most functions. \section{Citation} The following publications can be refered to learn more about the methods implemented in this package. Xu Z, Niu L, Taylor JA. Comprehensive evaluation of DNA methylation analysis pipelines for Illumina BeadChip. In review, 2021 Xu Z, Niu L, Li L, Taylor JA. ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip, Nucleic Acids Research, 2015 Niu L, Xu Z, Taylor JA. RCP: a novel probe design bias correction method for Illumina Methylation BeadChip. Bioinformatics, 2016 Xu Z, Langie SA, De Boever P, Taylor JA, Niu L. RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip. BMC Genomics. 2017 Xu Z, Taylor JA, Leung YK, Ho SM, Niu L. oxBS-MLE: an efficient method to estimate 5-methylcytosine and 5-hydroxymethylcytosine in paired bisulfite and oxidative bisulfite treated DNA. Bioinformatics. 2016 Xu Z, Taylor JA. Reliability of DNA methylation measures using Illumina methylation BeadChip. Epigenetics 2020 Xu Z, Xie C, Taylor JA, Niu L. ipDMR: Identification of differentially methylated regions with interval P values. Bioinformatics, 2020 Xu Z, Niu L, Taylor JA. Comprehensive evaluation of DNA methylation analysis pipelines for Illumina BeadChip in review \section{List of functions} \begin{tabular}{| l | l |} \toprule \textbf{Function} & \textbf{Description} \\ \midrule \textbf{Data acquisition} & \\ readidat & Read idat files into R \\ readmanifest & Read array manifest file into R \\ \midrule \textbf{Quality control} & \\ QCinfo & Extract and visualize QC information \\ plotCtrl & Generate internal control plots \\ getCGinfo & Extract CpG probe annotation information \\ calcdetP & Compute detection P values \\ qcfilter & Remove low quality values, samples or CpGs; \\ & remove outlier samples and perform imputation \\ QCfilter & Sample, CpG, low quality data filter \\ nmode & Identify gap probes \\ dupicc & Calculate ICC with data for duplicates \\ freqpoly & Frequency polygon plot for single variable \\ multifreqpoly & Frequency polygon plot for multiple variables \\ \midrule \textbf{Preprocessing} & \\ mpreprocess & Preprocessing pipeline \\ preprocessENmix & ENmix background correction and dye bias correction \\ relic & RELIC dye bias correction \\ norm.quantile & Quantile normalization \\ rcp & RCP probe design type bias correction \\ \midrule \textbf{Differential analysis} & \\ ipdmr & ipDMR differentially methylated region analysis \\ combp & Combp differentially methylated region analysis \\ \midrule \textbf{Other functions} & \\ oxBS.MLE & MLE estimates of 5-methylcytosine (5mC) and \\ & 5-hydroxymethylcytosine (5hmC) \\ estimateCellProp & Estimate cell type proportions \\ methyAge & Calculate methylation age \\ predSex & Estimate sample sex \\ ctrlsva & Derive surrogate variables to control for experimental \\ & confoundings using non-negative control probes \\ pcrplot & Principal component regression plot \\ mhtplot & P value manhattan plot \\ p.qqplot & P value Q-Q plot \\ B2M & Convert Beta value to M value \\ M2B & Convert M value to Beta value \\ \bottomrule \end{tabular} \section{Example Analysis} ENmix functions are easy to run. The following examples demonstrate how to perform a typical QC and data-preprocessing, as well as in depth methylation analyses. \subsection{Example 1: using pipeline} The pipeline function \Rfunction{mpreprocess} can be used to perform quality control and data preprocessing in just two sentences. <>= rgSet <- readidat(datapath) beta=mpreprocess(rgSet) @ These two sentences can achieve the following tasks: \begin{itemize} \item Read in raw idat files \item Data preprocessing \begin{itemize} \item background correction \item dye-bias correction \item between-array normalization \item probe-type bias adjustment \end{itemize} \item Quality control \begin{itemize} \item Identify and exclude low quality probes \item Identify and exclude low quality samples \item Identify and exclude outlier samples \item Remove low quality and outlier data points \item Imputation of missing values \end{itemize} \item Visualization \begin{itemize} \item Data distribution plots before and after low quality sample removal \item Plots of quality controls measures (detp, bisulfite) to guide selection of thresholds \end{itemize} \end{itemize} The following is an executable example. <>= library(ENmix) #read in data require(minfiData) #read in IDAT files path <- file.path(find.package("minfiData"),"extdata") rgSet <- readidat(path = path,recursive = TRUE) #data pre-processing beta=mpreprocess(rgSet,nCores=6) #quality control, data pre-processing and imputation beta=mpreprocess(rgSet,nCores=6,qc=TRUE,fqcfilter=TRUE, rmcr=TRUE,impute=TRUE) @ \subsection{Example 2: using individual function} The example code below is basically doing the same thing as in example 1, but has more customized options. <>= library(ENmix) #read in data path <- file.path(find.package("minfiData"),"extdata") rgSet <- readidat(path = path,recursive = TRUE) #QC info qc<-QCinfo(rgSet) #background correction and dye bias correction #if qc info object is provided, the low quality or outlier samples # and probes will be excluded before background correction mdat<-preprocessENmix(rgSet, bgParaEst="oob", dyeCorr="RELIC", QCinfo=qc, nCores=6) #between-array normalization mdat<-norm.quantile(mdat, method="quantile1") #probe-type bias adjustment beta<-rcp(mdat,qcscore=qc) #assign missing to low quality and outlier data points, remove samples # and probes with too many missing data, and do imputation beta <- qcfilter(beta,qcscore=qc,rmcr=TRUE,impute=TRUE) @ \subsection{Example 3: A more comprehensive example} This example is to demonstrate the flexibility of the package functions. Getting detailed information about data quality, SNP-like probes, excluding user-specified probes and/or samples, principal component regression analysis, low quality or outlier sample or probe removal, sex prediction, cell type proportion estimation, methylation age estimation, differentially methylated region analysis and surrogate variables, etc. <>= library(ENmix) #read in data path <- file.path(find.package("minfiData"),"extdata") rgSet <- readidat(path = path,recursive = TRUE) #attach some phenotype info for later use sheet <- read.metharray.sheet(file.path(find.package("minfiData"), "extdata"),pattern = "csv$") rownames(sheet)=paste(sheet$Slide,sheet$Array,sep="_") colData(rgSet)=as(sheet[colnames(rgSet),],"DataFrame") #generate internal control plots to inspect quality of experiments plotCtrl(rgSet) #generate quality control (QC) information and plots, #identify outlier samples in data distribution #if set detPtype="negative", recommand to set depPthre to <= 10E-6 #if set detPtype="oob", depPthre of 0.01 or 0.05 may be sufficient #see Heiss et al. PMID: 30678737 for details qc<-QCinfo(rgSet,detPtype="negative",detPthre=0.000001) ###data preprocessing #background correction and dye bias correction mdat<-preprocessENmix(rgSet, bgParaEst="oob", dyeCorr="RELIC", QCinfo=qc, nCores=6) #between-array normalization mdat<-norm.quantile(mdat, method="quantile1") #attach phenotype data for later use colData(mdat)=as(sheet[colnames(mdat),],"DataFrame") #probe-type bias adjustment beta<-rcp(mdat,qcscore=qc) #Search for multimodal CpGs, so called gap probes #beta matrix before qcfilter() step should be used for this purpose nmode<-nmode(beta, minN = 3, modedist=0.2, nCores = 6) #filter out poor quality and outlier data points for each probe; #rows and columns with too many missing values can be removed #Imputation will be performed to fill missing data if specify. beta <- qcfilter(beta,qcscore=qc,rmcr=TRUE, rthre=0.05, cthre=0.05,impute=TRUE) #If for epigenetic mutation analysis, use #beta <- qcfilter(beta,qcscore=qc,rmoutlier=FALSE,rmcr=TRUE, rthre=0.05, # cthre=0.05,impute=FALSE) ##Principal component regression analysis plot #phenotypes to be studied in the plot cov<-data.frame(group=colData(mdat)$Sample_Group, slide=factor(colData(mdat)$Slide)) #missing data is not allowed in the analysis! pcrplot(beta, cov, npc=6) #Non-negative control surrogate variables, which can be used in # downstream association analysis to control for batch effects. csva<-ctrlsva(rgSet) #estimate cell type proportions #rgDataset or methDataSet can also be used for the estimation celltype=estimateCellProp(userdata=beta, refdata="FlowSorted.Blood.450k", nonnegative = TRUE,normalize=TRUE) #predic sex based on rgDataSet or methDataset sex=predSex(rgSet) sex=predSex(mdat) #Methylation age calculation mage=methyAge(beta) #ICC analysis can be performed to study measurement reliability if # have duplicates, see manual #dupicc() #After association analysis, the P values can be used for DMR analysis #simulate a small example file in BED format dat=simubed() #using ipdmr method ipdmr(data=dat,seed=0.1) #using comb-p method combp(data=dat,seed=0.1) @ \section{Setting up the data} The first step is to import array raw data files (*.idat) to create an object of \Robject{rgDataSet}, which is similar to \Robject{RGChannelSetExtended} in minfi package, but with probe annotation integrated, and having smaller data size (about 1/3 smaller). Alternatively, User can also use minfi package to read in idat files and create \Robject{RGChannelSetExtended}. Most functions in ENmix support \Robject{RGChannelSetExtended} and \Robject{RGChannelSet}. Some array types and corresponding manifestfiles can be guessed by the program based on the number of probes per array. However, we recommand to provide correct manifest file directly, which can be easily downloaded from Illumina website, see below for some examples. This option also allows the function to read in data from newer array, such as mouse array. <>= library(ENmix) rgSet <- readidat(path = "directory path for idat files", recursive = TRUE) #Download manifestfile for HumanMethylation450 beadchip system("wget ftp://webdata2:webdata2@ussd-ftp.illumina.com/downloads/ProductFiles /HumanMethylation450/HumanMethylation450_15017482_v1-2.csv") mf="HumanMethylation450_15017482_v1-2.csv" rgSet <- readidat(path = "path to idat files",manifestfile=mf,recursive = TRUE) @ When methylation IDAT raw data files are not available, such as in many publicly available datasets, users can use methylated (M) and unmethylated (U) intensity data to create an object of \Robject{MethylSet}. \Robject{MethylSet} is also supported by most functions in ENmix. <>= M<-matrix_for_methylated_intensity U<-matrix_for_unmethylated_intensity pheno<-as.data.frame(cbind(colnames(M), colnames(M))) names(pheno)<-c("Basename","filenames") rownames(pheno)<-pheno$Basename pheno<-AnnotatedDataFrame(data=pheno) anno<-c("IlluminaHumanMethylation450k", "ilmn12.hg19") names(anno)<-c("array", "annotation") mdat<-MethylSet(Meth = M, Unmeth = U, annotation=anno, phenoData=pheno) @ \section{Quality Control} \subsection{Internal control probes} Both Illumina 450k and EPIC array incorporated 15 different types of internal control probes (total of 848 probes for 450K and 635 probes in EPIC). The control plots from ENmix function \Rfunction{plotCtrl} are similar to the control plots generated by Illumina GenomeStudio software. See Illumina Infinium HD Methylation Assay for detailed description on how to interpret these control figures. <>= plotCtrl(rgSet) @ Below is a list of internal control types and number of probes for each type. \begin{tabular}{ l c c} \toprule \textbf{Control types} & \textbf{450K} & \textbf{EPIC} \\ \midrule \textbf{Sample-Independent Controls} & & \\ STAINING & 4 & 6\\ EXTENSION & 4 & 4\\ HYBRIDIZATION & 3 &3 \\ TARGET REMOVAL & 2 & 2\\ RESTORATION & 1 & 1\\ \midrule \textbf{Sample-Dependent Controls} & & \\ BISULFITE CONVERSION I & 12 & 10\\ BISULFITE CONVERSION II & 4 & 4\\ SPECIFICITY I & 12 & 12\\ SPECIFICITY II & 3 & 3\\ NON-POLYMORPHIC & 4 & 9\\ NORM\_A & 32 & 27\\ NORM\_C & 61 & 58\\ NORM\_G & 32 & 27\\ NORM\_T & 61 & 58\\ NEGATIVE & 613 & 411\\ \bottomrule \end{tabular} \begin{figure}[!htbp] \centerline{\includegraphics[scale=0.75]{./fig/BISULFITE_CONVERSION_I.jpg}} \caption{Bisulfite conversion controls for type I probes} \label{fig:bisul_I} \end{figure} \begin{figure}[!htbp] \centerline{\includegraphics[scale=0.75]{./fig/NEGATIVE.jpg}} \caption{Negative control probes} \label{fig:negcont} \end{figure} \begin{figure}[!htbp] \centerline{\includegraphics[scale=0.75]{./fig/NORM_ACGT.jpg}} \caption{NORM ACGT control probes} \label{fig:acgt} \end{figure} These controls can be plotted in user specified order to check how experimental factors affect methylation measures, such as batch, plate, array or array location. <>= path <- file.path(find.package("minfiData"),"extdata") rgSet <- readidat(path = path,recursive = TRUE) sheet <- read.metharray.sheet(file.path(find.package("minfiData"), "extdata"), pattern = "csv$") rownames(sheet)=paste(sheet$Slide,sheet$Array,sep="_") colData(rgSet)=as(sheet[colnames(rgSet),],"DataFrame") #control plots IDorder=rownames(colData(rgSet))[order(colData(rgSet)$Slide, colData(rgSet)$Array)] plotCtrl(rgSet,IDorder) @ \subsection{Data distribution plots} Methylation intensity or beta value distribution plots are very useful for data summary, visual inspection and identification of outlier samples. Density plot is routinely generated using R function \Rfunction{multidensity}. However, the function is computationally intensive, and can take several hours to produce density plots for a large methylation dataset. Frequency polygon plot is an alternative figure for inspection of data distribution. Similar to histogram, it can accurately reflect data distribution and easy to understand. The function is also much faster, and only takes a few minutes to produce a distribution plot for >1000 samples. <>= mraw <- getmeth(rgSet) #total intensity plot is userful for data quality inspection #and identification of outlier samples multifreqpoly(assays(mraw)$Meth+assays(mraw)$Unmeth, xlab="Total intensity") #Compare frequency polygon plot and density plot beta<-getB(mraw) anno=rowData(mraw) beta1=beta[anno$Infinium_Design_Type=="I",] beta2=beta[anno$Infinium_Design_Type=="II",] library(geneplotter) jpeg("dist.jpg",height=900,width=600) par(mfrow=c(3,2)) multidensity(beta,main="Multidensity") multifreqpoly(beta,main="Multifreqpoly",xlab="Beta value") multidensity(beta1,main="Multidensity: Infinium I") multifreqpoly(beta1,main="Multifreqpoly: Infinium I", xlab="Beta value") multidensity(beta2,main="Multidensity: Infinium II") multifreqpoly(beta2,main="Multifreqpoly: Infinium II", xlab="Beta value") dev.off() @ See the following figures (Figure 4) generated from the above code. When type I and type II probes are plotted separately (Fig 4 bottom 4 panels) the difference in modes between type I and II probes can be appreciated. But when all probes are plotted together (Fig 4 top panels), the multidensity plot obscures these differences, while they remain readily apparent in the multifreqpoly plot. In addition, the multidensity plots appear to suggest that probes range in value from <0 to >1, whereas multifreqpoly correctly show the range from 0 to 1. \begin{figure}[!htbp] \centerline{\includegraphics[scale=0.75]{./fig/dist.jpg}} \caption{Methylation beta value distribution plots for all probes (top 2 panels) and for type I (middle panels) and II (bottom panels) probes separately. The smoothing function in multidensity plots (panels on left) results in misleading range and mode information which are more accurately depicted in the multifreqpoly plots (panels on right)} \label{fig:dist} \end{figure} \subsection{QC information, outlier samples, low quality samples and probes} Data quality measures, including detection P values, average intensities for bisulfite conversion probes and number of beads for each methylation read can be extracted or estimated using the function \Rfunction{QCinfo} from an object of \Robject{rgDataSet} or \Robject{RGChannelSetExtended}. Based on default or user specified quality score thresholds, the \Rfunction{QCinfo} will identify a list of low quality samples and CpG probes as wells as outlier samples based on total intensity or beta value distribution. These samples and probes should be excluded before further analysis. Data quality score figures "qc\_sample.jpg" and "qc\_CpG.jpg" from \Rfunction{QCinfo} can be used to guide the selection of quality score thresholds. Low quality samples and probes can be filtered out using \Rfunction{preprocessENmix} if the QCinfo object is provided. These samples and probes are listed as "badsample" and "badCpG" in QCinfo object, and also marked in red in figure "freqpolygon\_beta\_beforeQC.jpg". <>= qc<-QCinfo(rgSet) #exclude badsample and bad CpG before backgroud correction mdat<-preprocessENmix(rgSet, QCinfo=qc, nCores=6) @ \subsection{Filtering outliers, low quality data points, missing values and imputation} After excluding low quality samples and probes, as well as outlier samples. we can still have many low quality and outlier data points for many CpGs. These small percentage of data points can have big impact on downstream association statistical tests for individual CpGs. Function \Rfunction{qcfilter} can be used to filter out these data points and replace them with missing value NA. Outliers are defined as values smaller than 3 times IQR from the lower quartile or larger than 3 times IQR from the upper quartile. Some statistical methods do not allow missing values,(e.g. principal component analysis), so argument \textit{impute=TRUE} in the function can be specified to impute missing data using k-nearest neighbors method. <>= #filter out outliers only b1=qcfilter(beta) #filter out low quality and outlier values b2=qcfilter(beta,qcscore=qc) #filter out low quality and outlier values, remove rows #and columns with too many missing values b3=qcfilter(beta,qcscore=qc,rmcr=TRUE) #filter out low quality and outlier values, remove rows and #columns with too many (rthre=0.05,cthre=0.05, 5% in default) missing values, # and then do imputation b3=qcfilter(beta,qcscore=qc,rmcr=TRUE,rthre=0.05, cthre=0.05,impute=TRUE) @ \section{Background correction and dye-bias correction} Function \Rfunction{preprocessENmix} incorporates a model-based background correction method \textit{ENmix}, which models methylation signal intensities with a flexible exponential-normal mixture distribution, together with a truncated normal distribution to model background noise. Argument "dyeCorr" can be used to specify a method for dye-bias correction, the default is RELIC. See the following papers for the detailed description of related methods: Zongli Xu, et. al. ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip, Nucleic Acids Research, 2015 Xu Z, Langie SA, De Boever P, Taylor JA, Niu L. RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip. BMC Genomics. 2017 If argument QCinfo is specified, the low quality samples and probes identified by function \Rfunction{QCinfo} will be excluded before ENmix background correction. Using argument \textit{exSample} and \textit{exCpG}, User can also specify a list of samples or probes to be excluded before background correction. <>= qc=QCinfo(rgSet) mdat<-preprocessENmix(rgSet, bgParaEst="oob", dyeCorr="RELIC", QCinfo=qc, exCpG=NULL, nCores=6) @ \section{Inter-array normalization} Function \Rfunction{norm.quantile} can be used to perform quantile normalization on methylation intensity values. <>= mdat<-norm.quantile(mdat, method="quantile1") @ \section{Probe-type bias adjustment} The majority of probes on Illumina 450K and EPIC BeadChips are type II probes. Although type II probes facilitate increased array genome coverage, they were shown to have decreased dynamic range and reproducibility compared to type I probes. Taking advantage of the high spatial correlation of DNA methylation levels along the human genome, The RCP (Regression on Correlated Probes) method utilizes nearby (<25 bp) type I and II probe pairs to derive the quantitative relationship between probe types and then recalibrates type II probe measurements using type I probes as referents. <>= beta<-rcp(mdat) @ See the following publication for the detailed description of the method: Niu L, Xu Z, Taylor JA. RCP: a novel probe design bias correction method for Illumina Methylation BeadChip. Bioinformatics, 2016 \section{Batch effect correction} Function \Rfunction{ctrlsva} can be used to estimate surrogate variables for batch effects and unknown experimental confounders using intensity data for non-negative internal control probes. These surrogate variables can then be modeled as covariables in downstream association analysis to adjust for experimental variation. <>= sva<-ctrlsva(rgSet) @ \section{Principal component regression analysis plot} First, principal component analysis will be performed on the standardized beta value matrix (standardized probe by probe), and then the specified number of top principal components (that explain most data variation) will be used to perform linear regression with each specified variables, such as batch or phenotype variables. Regression P values will be plotted to explore methylation data variance structure and to identify possible confounding variables to guide association statistical analysis. Principal components are ordered according to the percentage of variance they explained from large to small. PCA will not allow for missing values, so specify impute=true in preprocessing if there is missing data. <>= require(minfiData) mdat <- preprocessRaw(RGsetEx) beta=getBeta(mdat, "Illumina") group=pData(mdat)$Sample_Group slide=factor(pData(mdat)$Slide) cov=data.frame(group,slide) pcrplot(beta,cov,npc=6) @ \begin{figure}[!htbp] \centerline{\includegraphics[scale=0.75]{./fig/pcr_diag.jpg}} \caption{Example principal component regression p value plot of raw data generated using 450K methylation data from a published study} \label{fig:pcrdiag} \end{figure} \section{Multimodal CpGs or gap probes} Function \Rfunction{nmode} uses an empirical approach to identify CpGs whose methylation values are multimodal distributed (SNP-like or gap probes). When measured in a population of people the majority of CpGs on the Illumina HumanMethylation450 BeadChip have unimodal distributions of DNA methylation values with relatively small between-person variation. However, some CpGs may have multimodal distributions of methylation values with sizeable differences between modes and large between-person variation. These multimodal distributed data are usually caused by SNP effect in the probe region, problematic probe design or other unknown artifacts rather than by actual difference in methylation level and thus should be excluded from DNA methylation analysis. Researchers have often excluded CpGs based simply on SNP annotation information (e.g. reported SNPs in or near probe sequences). However, because SNP frequence and annotation always depends on population origin, we find that this approach alone may exclude many well-distributed (unimodal) CpGs, while still failing to identify other problematic multi-modal CpGs. We developed an empirical approach to identify CpGs that are obviously not uni-modally distributed, so that researchers can make more informed decisions about whether to exclude them in their particular study populations and analyses. See online supplementary materials of the following paper for an evaluation of the method using published EWAS data. Zongli Xu, Liang Niu, Leping Li and Jack A. Taylor ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip, Nucleic Acids Research, 2015 <>= nmode<- nmode(beta, minN = 3, modedist=0.2, nCores = 5) @ \section{Cell type proportion estimation} Whole blood samples are often used in EWAS, however, whole blood has mixed cell types and methylation level can be widely different between different cell types. Function \Rfunction{estimateCellProp} can be used to estimate cell types based on reference dataset. Users should select a reference that is most resemble to their own study samples. Currently there are 6 publicly deposited reference datasets: "FlowSorted.Blood.450k", "FlowSorted.DLPFC.450k","FlowSorted.CordBlood.450k", "FlowSorted.CordBloodCombined.450k", "FlowSorted.CordBloodNorway.450k" or "FlowSorted.Blood.EPIC". <>= require(minfidata) path <- file.path(find.package("minfiData"),"extdata") #based on rgDataset rgSet <- readidat(path = path,recursive = TRUE) celltype=estimateCellProp(userdata=rgSet, refdata="FlowSorted.Blood.450k", nonnegative = TRUE,normalize=TRUE) @ The input data can be in one of the follow format: \Rfunction{rgDataSet}, \Rfunction{methDataSet}, \Rfunction{MethylSet}, \Rfunction{RGChannelSet} or methylation beta value matrix. \section{Methylation age estimation} Methylation age can reflect a person's biological age, which may be more related to the person's health status than chronological age. Three different types of methylation age can be estimated using \Rfunction{methyAge}: Horvath, Hannum, PhenoAge and DunedinPACE. <>= require(minfidata) path <- file.path(find.package("minfiData"),"extdata") #based on rgDataset rgSet <- readidat(path = path,recursive = TRUE) meth=getmeth(rgSet) beta=getB(meth) mage=methyAge(beta) @ \section{Differentially methylated regions (DMRs)} EWAS is usually done probe by probe. However, in the human genome, nearby CpGs often have similar methylation status, so combined statistics from nearby CpGs can potentially increase association signal. Functions \Rfunction{ipdmr} and \Rfunction{combp} can utilize P values from EWAS to identify differentially methylated regions (DMRs). The argument "seed" in these two functions indicate the FDR threshold for initial selection of DMR regions, but we note that same value of seed do not appear to have the same effect in these two functions. It is often more stringent in ipdmr than combp. There is no true DMR in the following example data, we set less strengent seed threshold is only for demostration purpose to show what the output looks like when DMRs were detected. <>= dat=simubed() names(dat) ipdmr(data=dat,seed=0.1) combp(data=dat,seed=0.1) @ \section{Intraclass correlation coefficient (ICC) reliability measures} Several studies have showen that a large percentage of CpGs on Illumina arrays have poor reliability, i.e. they have low correlation between replicate measures in a same set of samples. Measurement of the reliability of individual CpGs can be assessed by calculating intraclass correlation coefficients (ICC) using methylation from duplicate samples. Function \Rfunction{dupicc} can be used for this purpose. <>= require(minfiData) path <- file.path(find.package("minfiData"),"extdata") rgSet <- readidat(path = path,recursive = TRUE) mdat=getmeth(rgSet) beta=getB(mdat,"Illumina") #assuming list in id1 are corresponding duplicates of id2 dupidx=data.frame(id1=c("5723646052_R02C02","5723646052_R04C01", "5723646052_R05C02"), id2=c("5723646053_R04C02","5723646053_R05C02", "5723646053_R06C02")) iccresu<-dupicc(dat=beta,dupid=dupidx) @ \section{5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC)} Studies showed that 5-Methylcytosine and 5-hydroxymethylcytosine can have very different function. However, traditional bisulfite (BS) treatment commonly used in genome-wide methylation studies can not distinguish 5mC from 5hmC. Sequencing or array data from paired DNA samples (bisulfite and oxidative bisulfite conversion) are needed for the separate estimation. Bisulfite converted DNA can be used to estimate the proportion of 5mC+5hmC and oxidative bisulfite (oxBS) converted DNA can be used to estimate the proportion of 5mC. Function \Rfunction{oxBS.MLE} can be used to generate the maximum Likelihood Estimate (MLE) of 5mC and 5hmC using sequencing or array data from paired experiments (Xu et al. Bioinformatics 2016). <>= load(system.file("oxBS.MLE.RData",package="ENmix")) resu<-oxBS.MLE(beta.BS,beta.oxBS,N.BS,N.oxBS) dim(resu[["5mC"]]) dim(resu[["5hmC"]]) @ \section{Compatibility with other related R packages} Most functions in the \Rpackage{ENmix} support the data structure provided by \Rpackage{minfi} package. The same data structures were also used by several other R packages, such as \Rpackage{ChAMP} and \Rpackage{wateRmelon}, so the output from ENmix functions can be easily utilized in these packages for further analysis. Here are some examples: Example 1: mixed use of minfi and ENmix functions <>= library(ENmix) #minfi functions to read in data sheet <- read.metharray.sheet(file.path(find.package("minfiData"), "extdata"), pattern = "csv$") rgSet <- read.metharray.exp(targets = sheet, extended = TRUE) #ENmix function for control plot plotCtrl(rgSet) #minfi functions to extract methylation and annotation data mraw <- preprocessRaw(rgSet) beta<-getB(mraw, "Illumina") anno=getAnnotation(rgSet) #ENmix function for fast and accurate distribution plot multifreqpoly(beta,main="Data distribution") multifreqpoly(beta[anno$Type=="I",],main="Data distribution, type I") multifreqpoly(beta[anno$Type=="II",],main="Data distribution, type II") #ENmix background correction mset<-preprocessENmix(rgSet, bgParaEst="oob", dyeCorr="RELIC", nCores=6) #minfi functions for further preprocessing and analysis gmSet <- preprocessQuantile(mset) bumps <- bumphunter(gmSet, design = model.matrix(~ gmSet$status), B = 0, type = "Beta", cutoff = 0.25) @ \section{SessionInfo} <>= toLatex(sessionInfo()) @ \section{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 Liang Niu, Zongli Xu and Jack A. Taylor, RCP: a novel probe design bias correction method for Illumina Methylation BeadChip. Bioinformatics 2016 Zongli Xu, Jack A. Taylor, Yuet-Kin Leung, Shuk-Mei Ho and Liang Niu, oxBS-MLE: An efficient method to estimate 5-methylcytosine and 5-hydroxymethylcytosine in paired bisulfite and oxidative bisulfite treated DNA. Bioinformatics 2016 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 Zongli Xu, Jack A Taylor. Reliability of DNA methylation measures using Illumina methylation BeadChip. Epigenetics 2020 Zongli Xu, Changchun Xie, Jack A. Taylor, Liang Niu, ipDMR: Identification of differentially methyl-ated regions with interval p-values. Bioinfomatics 2020 Illumina Inc., Infinium HD Assay Methylation Protocol Guide, Illumina, Inc. San Diego, CA. 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. 13631369. Horvath S. DNA methylation age of human tissues and cell types. Genome biology 2013 14:R115. Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Molecular cell 2013 49:359-367. Levine ME, Lu AT, Quach A, Chen BH, Assimes TL, Bandinelli S, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY) 2018 10:573-591. Pedersen BS1, Schwartz DA, Yang IV, Kechris KJ. Comb-p: software for combining, analyzing, grouping and correcting spatially correlated P-values. Bioinfomatics 2012 EA Houseman, WP Accomando, DC Koestler, BC Christensen, CJ Marsit, HH Nelson, JK Wiencke and KT Kelsey. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC bioinformatics (2012) 13:86. Heiss JA, Allan C Just. Improved filtering of DNA methylation microarray data by detection p values and its impact on downstream analyses. Clinical Epigenetics, 2019 Xu Z, Niu L, Taylor JA. The ENmix DNA methylation analysis pipeline for Illumina BeadChip and comparisons to 7 other preprocessing pipelines. Clinical Epigenetics, 2022 Daniel W Belsky, Avshalom Caspi, David L Corcoran,et al. DunedinPACE, a DNA methylation biomarker of the pace of aging, eLIfe, 2022 \end{document}