methylkey: DNA Methylation Analysis from Illumina Arrays

methylkey is a Bioconductor package for comprehensive analysis of DNA methylation data from Illumina methylation arrays. It supports multiple platforms (27k, 450k, EPIC, EPIC+, EPICv2, and mouse arrays) with preprocessing pipelines using both sesame and minfi packages.

Features

• Flexible Preprocessing: Choose between sesame or minfi pipelines
• Quality Control: Comprehensive QC plots and statistics
• Differential Methylation: DMP and DMR detection using multiple methods
• Platform Detection: Automatic detection of array platform
• Sex Inference: Automatic inference of sample sex
• Epigenetic Age: Multiple clock model support
• Cell Type Estimation: Blood immune cell composition estimation
• Standardized Output: SummarizedExperiment-based objects for downstream analysis

Installation

In Rstudio

install.packages("pak")

pak::pkg_install(c(
  "cran::ggsankeyfier", 
  "github::perishky/dmrff",
  "github::sailalithabollepalli/EpiSmokEr",
  "github::chiaraherzog/IlluminaMouseMethylationanno.12.v1.mm10",
  "github::IARCbioinfo/methylkey"
))

With conda environment :

1. Create the conda environment

mamba env create -f environment.yml
conda activate methylkey

2. Install other libraries and methylkey in R

library(pak)
pak::pkg_install(c("cran::ggsankeyfier", "github::perishky/dmrff"))
pak::pkg_install("github::sailalithabollepalli/EpiSmokEr")
pak::pkg_install("chiaraherzog/IlluminaMouseMethylationanno.12.v1.mm10")
pak::pkg_install("github::IARCbioinfo/methylkey")

Requirements

• R >= 4.2.1
• Bioconductor packages (automatically installed)
• For minfi pipeline: minfi, wateRmelon
• For sesame pipeline: sesame, sesameData

Quick Start

1. Prepare Sample Sheet

Create a tab-separated file with sample information:

Sample_Name  Basename                  Group   Age
Sample_001   203021070069_R03C01       Ctrl    45
Sample_002   203021070069_R04C01       Case    50
Sample_003   203021070069_R05C01       Ctrl    48

Required columns:

• Basename or barcode: Sentrix ID and position (e.g., 203021070069_R03C01)
• At least one additional column for grouping or clinical variables

Load the sample sheet in R:

sampleSheet <- readr::read_delim("path/to/sampleSheet.txt")

2. Preprocessing with sesame

For mouse arrays or when sesame is preferred:

library(methylkey)

# Load IDAT files and preprocess with sesame
meth <- sesame2Betas(
  idat = "path/to/idat/directory",
  sampleSheet = sampleSheet,
  prep = "QCDPB",
  na = 0.2,
  ncore = 4
)

# View object information
meth
colData(meth)
get_betas(meth)[1:5, 1:5]

3. Preprocessing with minfi

For 450k and EPIC arrays:

meth <- minfi2Betas(
  idat = "path/to/idat/directory",
  sampleSheet = sampleSheet,
  na = 0.2,
  compositeCellType = "Blood"  # Optional: estimate cell counts
)

4. Quality Control

# Access QC statistics
qcs <- metadata(meth)$qcs
head(qcs)

# Generate QC plots
plot_Detection(meth)
plot_NA(meth)

5. Differential Methylation Analysis

# Convert to M-values and perform analysis
mvals <- get_mvals(meth, grp = "group")

# Run differential methylation analysis
mrs <- methyldiff(
  se = meth,
  model = "~group",
  intercept = "Ctrl",
  method = "ls",
  fdr = 0.05,
  pcutoff = 0.2
)

# Extract results
dmps <- get_dmps(mrs)
dmrs <- get_dmrs(mrs)

Supported Array Platforms

Platform	Probes	Support
IlluminaHumanMethylation27k	27,578	✓
IlluminaHumanMethylation450k	486,427	✓
IlluminaHumanMethylationEPIC	866,553	✓
IlluminaHumanMethylationEPIC+	865,546	✓
IlluminaHumanMethylationEPICv2	937,690	✓
IlluminaMouseMethylation285k	296,070	✓

Data Objects

Betas Object

Container for beta values (0-1 scale):

# Access
betas <- get_betas(meth)
samples <- colData(meth)
platform <- metadata(meth)$plateform

Mvals Object

Container for M-values (log2 scale) for statistical analysis:

mvals <- get_mvals(betas)
mvals_matrix <- assays(mvals)$mvals

Documentation

• Full function documentation: ?methylkey or help(package="methylkey")
• Vignettes: browseVignettes("methylkey")
• Examples: See examples/ directory for complete workflows

Workflow Summary

IDAT Files
    ↓
[sesame2Betas or minfi2Betas]
    ↓
Betas Object
    ↓
[QC & Visualization]
    ↓
[get_mvals]
    ↓
Mvals Object
    ↓
[methyldiff]
    ↓
MethylResultSet (DMPs & DMRs)

Common Issues

"Package sesame not found"

BiocManager::install("sesame")

"Basename not in sampleSheet" Ensure your sample sheet has a Basename column with barcode information.

"Barcodes in sampleSheet not found in betas" Verify that barcodes in the sample sheet match IDAT file names exactly.

Citation

If you use methylkey in your research, please cite:

@software{methylkey2026,
  author = {Cahais, Vincent},
  year = {2026},
  title = {methylkey: DNA methylation analysis from Illumina arrays},
  url = {https://github.com/IARCbioinfo/methylkey}
}

References

• Aryee MJ, et al. (2014). Minfi: A flexible and comprehensive Bioconductor package for the analysis of Infinium DNA Methylation microarrays. Bioinformatics, 30(10), 1363–1369.
• Zhou W, et al. (2018). SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions. NAR, 46(20), e119.
• Hansen KD, et al. (2016). Reference-based correction of interstrand dye bias in two-color microarrays. Genome Biology, 17(1), 37.

License

GPL-3.0

Contributing

Contributions are welcome! Please open an issue or pull request on GitHub.

Contact

For questions and issues: GitHub Issues