SeSAMe is an R package for processing Infinium DNA methylation data. SeSAMe currently supports EPICv2, EPIC, HM450, HM27, MSA, MM285 and Mammal40 platforms and dynamically generated manifest.
To install from Github,
BiocManager::install("zwdzwd/sesame")See the package Home Page on Bioconductor and the Developmental Branch.
It also has a depended data package for annotation and example data.
SeSAMe comes as two implementations that share the same methods and are developed together:
-
sesame (this package) — the reference implementation of the SeSAMe methods and the interactive R toolkit for exploratory analysis and KYCG enrichment. Actively maintained on Bioconductor, and the oracle the C version is validated against.
-
sesame-cli — a standalone C program (no R, no Bioconductor) covering the full workflow: IDAT → betas → QC → differential methylation → copy number, with visualization through cinderplot. Built for large cohorts, pipeline deployment, and non-R environments. A good fit when you process many samples or run in a pipeline.
conda install -c zhou-lab -c conda-forge sesame-cli
The two agree closely but are not bit-exact: sesame-cli reimplements the preprocessing independently, and where the R code has a numerically fragile construction it is reformulated. On probes with identical raw input, betas agree to a median of ~6e-6 and ~1e-3 worst case. Every intentional divergence is justified and recorded in the sesame-cli NUMERICS.md.
Bug reports are appreciated. Register issues at the SeSAMe issue tracker.
Please cite and reference SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions for more details.