SEnsible Step-wise Analysis of DNA MEthylation

latest sesame2 ↗ Cite ↗

Infinium analysis made simpler — the same betas as the R package, ~20× faster.

SeSAMe2 (2.x, this command line) and sesame (1.x, the R/Bioconductor package) are parallel, not sequential — one method, two implementations. R: BiocManager::install("sesame"); shell: conda install -c zhou-lab sesame.

GitHub· R / Bioconductor· R docs· YAME· InfiniumAnnotation· cinderplot· EPIC · EPICv2 · HM450 · MSA

Install & fetch — then fully offline

# the toolchain (sesame + yame is enough to run)
conda install -c zhou-lab -c conda-forge \
  sesame yame cinderplot tabl

# or build sesame from source
git clone --recurse-submodules \
  https://github.com/zhou-lab/sesame
cd sesame && make

# one-time: annotation into the shared store
# -y is required off a terminal; it refuses without
yame fetch -y EPICv2
yame fetch -y hg38

# a name is that directory's own files. -R adds everything
# beneath it (hg38/KYCG, hg38/models: 5.2 GB, mostly not
# sesame's). -n -R hg38 shows what that reaches first.

…builds a local, digest-verified store ($YAME_DATA_HOME, default ~/.local/share/yame) — offline from here on:

EPICv2/  # probe annotation
├ EPICv2.ordering.tsv.gz
├ EPICv2.hg38.mask.cm(+.idx)
├ EPICv2.hg38.coord.tsv.gz
├ EPICv2.hg38.snp.tsv.gz
├ EPICv2.typeI_ext.tsv.gz
└ SHA256SUMS
hg38/  # genome tracks
├ seqinfo.tsv.gz  gaps.tsv.gz
├ cytoband.tsv.gz  genes.bed.gz(+.tbi)
└ SHA256SUMS

Preprocess — IDAT → betas

QCDPB (openSesame default) over a cohort in one process → YAME .cg + qc.tsv.

# a folder of IDATs, searched recursively
sesame preprocess --out out/ in/

# or explicit prefixes (just betas)
sesame preprocess --output beta --out out/ \
  idat_prefix_1 idat_prefix_2

…writes one .cg per output over the cohort (each with a .cg.idx of sample names) + qc.tsv:

out/
├ beta.cg (+.idx)          # betas (fmt4)
├ intensity.cg (+.idx)     # M/U (fmt3)
├ pval.cg (+.idx)          # detection p
└ qc.tsv                   # QC metrics

qc.tsv holds 66 per-sample metrics. frac_dt — the fraction of probes above detection — is the most critical for assessing data quality (also per type, frac_dt_cg/_ch/_rs).

# head qc.tsv  (key columns)
sample    frac_dt  frac_dt_cg  frac_dt_rs
sampleA   0.978    0.979       1.00
sampleB   0.994    0.995       1.00

More knobs: --prep QCEPB swaps in linear dye bias (E = dyeBiasL); --collapse averages EPICv2/MSA replicate probes to their cg-prefix (betasCollapseToPfx); --detection pneg emits the negative-control-ECDF detection p; and GCT (bisulfite conversion) joins qc.tsv on every platform.

Bimodal distribution of GM12878 beta values — GM12878 betas — the classic bimodal shape: unmethylated (~0) and methylated (~1).

Read a .cg back — describe-probe

Betas and signals are stored positionally — one value per probe, in a fixed order, with no Probe IDs inside. The IDs live once in the platform's ordering table, which is what keeps the files small and byte-for-byte alignable: stack N samples into a matrix by column with no key-matching. describe-probe joins the two back into a labelled TSV.

$ sesame describe-probe out/beta.cg
Probe_ID          206909630040_R03C01
cg00000029_TC21   0.9124
cg00000769_BC11   0.0331
# the ordering is inferred from the row count -- each
# platform's is a distinct length. Name it if you'd rather:
$ sesame describe-probe --platform EPICv2 out/beta.cg
# --all for every sample column (matrix)
# the other direction: probe IDs -> coordinates, for yame rowsub -L
$ sesame describe-probe --platform HM450 --genome hg38 probes.txt
cg00000029        chr16_53434200

Peek at a raw IDAT — idat-dump

The bead-level record as Illumina wrote it, before any of it becomes a probe.

$ sesame idat-dump --tsv sample_Grn.idat.gz
# rows only: addr, mean, sd, nbeads. The header
# belongs to the summary form (no --tsv).
10600313   4821  118  12
10600322   287   34   9

Array QC plots — preprocess → cinderplot

The two diagnostics from sesame's supplemental QC, with no R in the loop: preprocess writes the signals, describe-probe labels them, tabl reshapes them, and cinderplot draws them.

Dye bias Q-Q plot, Infinium-I red versus green signal — **Dye bias Q-Q** — green quantiles below the diagonal: red reads hotter across the range. Re-run after `--prep CD` and the curve should land on the line. Ports `sesameQC_plotRedGrnQQ()`.

Total intensity versus beta, as a binned density — **Intensity vs beta** — two lobes at beta ~0.05 and ~0.9, funnelling to 0.5 as signal dies into background below ~2¹⁰. That funnel is where a detection cutoff belongs. Ports `sesameQC_plotIntensVsBetas()`.

Differential methylation — dml

Per-probe OLS with t/F tests + BH, matching R's lm to ~1e-9.

dml is the one command that takes no --platform: a .cg is positional, so it needs the ordering itself, by path ($STORE = $YAME_DATA_HOME). It still prints a deprecation notice for --index, which is the only flag it has — the rename has not reached it.

sesame dml --betas out/beta.cg \
  --index $STORE/EPICv2/EPICv2.ordering.tsv.gz \
  --meta samples.tsv \
  --formula '~ group + age' > dml.tsv

De-identify — deidentify

Strip the genetic fingerprint: zero (or reversibly scramble) the SNP (rs) probe intensities in an IDAT — every other bead is byte-identical, so betas are unchanged except the SNPs.

# zero SNP beads (a prefix does both channels)
sesame deidentify sample_prefix

# or reversibly scramble, then restore
sesame deidentify --randomize --seed 42 s.idat
sesame deidentify -r --seed 42 s_noid.idat

Copy number — cnv

OLS vs a normal panel → log2 → genome bins → CBS (DNAcopy port). Two output files.

sesame preprocess --prep "" --raw-signal \
  --output total_intensity --out t/ tumor
sesame cnv --platform EPICv2 \
  t/total_intensity.cg segments.tsv bins.tsv
# --probes: per-probe; --exclude chrY

The normal panel ships with the annotation (<P>.cnvnormals.cg, v8.2+), so yame fetch brings it; --normals points at your own instead.

cnv">

# plot the two files with cinderplot
cinderplot 'out/bins.tsv
  + aes(chrom=chrom, x=start, xend=end, y=log2ratio, colour=log2ratio)
  + geom_point(size=.5)
  + geom_segment(data="out/segments.tsv", y=seg.mean, color="black")
  + scale_x_genome("hg38/seqinfo.tsv.gz")
  + scale_colour_gradient2(low="#762a83", high="#b2182b")
  + ideogram("hg38/cytoband.tsv.gz")' cnv_k562.png

K562 (EPICv2/hg38): chr9p deletion, chr13/22 amps.

Genotype SNPs → VCF — vcf

SNP (rs) probes → VCF (formatVCF); GT + fractions exact vs R.

sesame vcf tumor --platform EPICv2 > geno.vcf
# the SNP table comes from the store. --snp
# overrides it, and is a path -- a bare
# filename is not looked up in the store.
# --variants: only rs + channel-switch

Region & gene views — region → cinderplot

A locus's betas as long-form TSV → cinderplot over the gene models in zhou-lab/genomes.

DNMT3A methylation on a probe-anchored heatmap — **--gene DNMT3A** — gene models + probe-anchored heatmap.

HOXA cluster methylation on a probe-anchored heatmap — **chr:beg-end** — the HOXA cluster.

Where we differ from R — DIVERGENCES.md

R is the oracle, and nearly everything matches it to the ULP. The exceptions are listed — not rounded away — in DIVERGENCES.md, each with what differs, why, and the measurement that settled it.

Most of them are not arithmetic at all: the two programs read different annotation. sesame reads the store (InfiniumAnnotation v8.1); R reads the copy inside sesameData. Where those disagree — 22 EPICv2 SNP probes whose colour channel differs, 635 EPIC control probes named ctl_10609447_NEGATIVE here and ctl_10609447 there — the newer annotation is the one this build is correct against.

# the tests grade against frozen values, not a live R
tests/truth/vcf.EPICv2.tsv.gz   # 127,572 rows, 22 C-sourced
tests/truth/PROVENANCE.tsv      # R, sesame, sesameData, store

# regenerating refuses to freeze an unexplained value
make truth RSCRIPT=Rscript-4.6.0

Speed — vs the R package

One EPICv2 sample, IDAT → betas, median of 3. R is timed twice: the openSesame() call alone, and the whole invocation — startup and library(sesame) included — which is what a pipeline pays, since it starts a fresh R per sample.

Runtime comparison: sesame C versus R openSesame — **2.5 s** vs **30.7 s** in-session (**12×**) and **46.9 s** per invocation (**19×**). One machine, one array — regenerate rather than quoting these.

Betas utilities — mliftover · impute

Post-process a beta.cg → beta.cg, positional to the platform ordering.

# lift betas across platforms (mLiftOver)
sesame mliftover --to EPIC --platform EPICv2 \
  beta.cg beta.EPIC.cg

# into a genome's CpG universe, as M/U pseudo-counts --
# what a whole-genome model (methscope classify) reads
sesame mliftover --to hg38 --platform MSA \
  --simulated-depth 100 beta.cg beta.hg38.cg

# a whole-genome model's mask down to the array
sesame mliftover --to MSA --platform hg38 \
  model.mrmp model.MSA.mrmp

# fill NA betas (mean / genomic neighbours)
sesame impute --method mean beta.cg out.cg
sesame impute --method neighbors \
  --platform EPICv2 beta.cg out.cg

mliftover is one index-to-index map applied one way: array→array joins on probe-ID prefix (EPICv2↔EPIC↔HM450); array↔genome joins on coordinate, through the platform's coord table and the genome's cpg_nocontig.cr, built in memory each run (~2 s, nothing to fetch). Only cg probes cross, decided by ID — an rs or nv probe carries genotype, not methylation, wherever it maps; replicates on one CpG are averaged; cg probes with no CpG row are dropped; every count is reported. The way back is lossy — CpGs under no probe are gone. Any format lifts, a .cm mask included; --threads N runs the records across workers that share the one map. impute fills NAs by the probe's cross-sample mean, or the nearest same-strand genomic neighbours (imputeBetas*).

Both are local fills — a neighbouring CpG or the column mean — and they only reach probes the array already carries. For imputation as an inference problem, use methscope: methscope upscale predicts genome-wide CpG methylation from a sparse methylome with a trained model, which is both more accurate and not limited to the manifest.

A platform is its ordering table — --platform takes one by path as readily as by name.

The ordering table — IDs live here, once

Probe_ID           M         U         col
cg00000029_TC21    NA        41791408  2
cg00000769_BC11    26756338  59799309  G
ctl_10609447_NEGATIVE  NA  10609447  2

Four columns and nothing else — not the Illumina/sesame manifest with tabs instead of commas. M is the methylated bead address (Illumina AddressB_ID, NA for Infinium-II), U the unmethylated one (AddressA_ID), and col the design: G / R Infinium-I in green or red, 2 Infinium-II. Sequences and coordinates live in separate files a custom array can do without. A legacy fifth mask column is still read; published orderings keep the mask in a companion .cm.

Tab- or comma-separated, plain or gzipped — the separator is sniffed from the header, so a table exported straight out of a CSV manifest needs no conversion. A header line is required.

Row order is load-bearing and permanent. A .cg stores one value per row with no Probe_IDs inside, so row i is probe i forever — sorting or inserting a row re-labels every .cg written against the old file. Freeze it when you publish, and version it if it changes.

Your own array — hand it the ordering

--platform takes a name or a path, so a custom or pre-release array needs no special flag. Every companion asset has its own override, inferred only when the platform is a name you published.

# named platform: ordering, mask, coords all inferred
sesame preprocess --platform EPICv2 --out out/ idats/

# custom array: the ordering IS the platform
sesame preprocess --platform MyArray.ordering.csv \
  --mask MyArray.mask.cm --prep QCDPB --out out/ idats/

# no mask: prep C or "", and betas only -- the
# default --output carries pval and qc, which
# need a mask whatever the prep
sesame preprocess --platform MyArray.ordering.csv \
  --prep C --output beta --out out/ idats/

Overrides: --mask --coords --snp --normals. From an Illumina manifest the ordering is a projection: IlmnID→Probe_ID, AddressB_ID→M, AddressA_ID→U, design + colour channel→col. The four columns are specified above.

Making the mask — a .cm bitset

A .cm is a bitset with one bit per ordering row, in the same order — 1 means drop the probe. Write the column, pack it, name the track:

# one 0/1 per ordering row, in the ordering's order
# replace the condition -- an undefined name is
# 0 to awk, so this masks nothing at all
awk -F'\t' 'NR>1{print ($1 ~ /^ctl_/) ? 1 : 0}' \
  MyArray.ordering.csv > m.txt

yame pack -f b m.txt MyArray.mask.cm   # b = bitset
echo M_custom > names.txt
yame index -s names.txt MyArray.mask.cm  # name it

Given directly, --mask wants a single-track .cm — a published platform's mask holds many (EPICv2 has 25) and sesame picks the recommended subset by platform name, so handing one over without a name would be ambiguous. It says so rather than guessing.

Reference — every subcommand's `-h`, as this build prints it

Generated from the binary at build time, so nothing here can drift from what sesame <command> -h says. Each entry is folded — click to open. In every example on the other tabs the highlighted subcommand links to its entry here and opens it. Grouped as the sesame help banner groups them.

Preprocessing

sesame preprocess — IDATs -> betas / signal / detection p-value / QC (QCDPB)


Usage:
  sesame preprocess [options] <prefix|dir> [<prefix|dir> ...]

  Preprocess Infinium IDATs to methylation levels. Each argument is an
  IDAT prefix (<prefix>_Grn.idat / _Red.idat, .gz ok) OR a directory,
  searched recursively for pairs (found prefixes are sorted). Applies
  --prep to every sample and writes one indexed YAME .cg per requested
  output, over the whole cohort. A failed sample becomes an NA column
  and the exit status is 1.

Options:
  --output LIST     comma list (default beta,intensity,pval,qc) from
                    beta, intensity, total_intensity, pval, qc
  --prep STEPS      prep code (default QCDPB): Q qualityMask,
                    C inferChannel, D dyeBias(NL), E dyeBias(linear),
                    P pOOBAH, B noob. Empty string = raw betas.
  --raw-signal      take intensity/total from the raw signal, not the prep
  --detection M     pval source: poobah (default) or pneg (negative-
                    control ECDF). Masking always uses pOOBAH.
  --collapse        average replicate probes to their cg-prefix in the
                    beta output; axis -> beta.cg.probes
  --platform P      EPIC | EPICv2 | HM450 | MSA (else from the bead
                    count), or the path to an ordering file, which is
                    how a custom array is named.
  --mask FILE       quality/background mask .cm. Inferred for a named
                    platform; required for a custom array using Q/P/B.
  --min-beads N     mask probes with < N beads (default 0)
  --out DIR         output directory (default: current directory)
  --tmp DIR         scratch for the sample-major matrix (default $TMPDIR)
  --threads N, -t   worker threads (default: number of CPUs)

Outputs (under --out):
  beta.cg  fmt4 betas     intensity.cg  fmt3 M/U
  pval.cg  fmt4 det. p    total_intensity.cg  fmt4
  qc.tsv   per-sample QC

Notes:
  Without --index the platform comes from the bead count and the
  ordering from the shared store ($YAME_DATA_HOME), then ./.

Analysis

sesame dml — differential methylation: per-probe linear models


Usage:
  sesame dml --betas <beta.cg|matrix.tsv> [options]
  (--formula '~a+b' --meta <s.tsv> | --design <X.tsv>)

  Per-probe differential methylation: OLS of each probe's betas on the
  design, with per-coefficient t-tests, a holdout F-test per
  categorical variable, effect sizes and BH-adjusted p-values. TSV on
  stdout, one row per probe.

Options:
  --betas FILE      a preprocess beta.cg (needs --index for Probe_IDs)
                    or a Probe_ID x sample matrix TSV
  --formula '~a+b'  model over --meta columns (categoricals dummied)
  --meta FILE       sample table; first column matches sample names
  --design FILE     numeric design matrix (for interactions), in
                    place of --formula/--meta
  --threads N, -t   worker threads (default: number of CPUs)

sesame cnv — copy number: log2 ratio vs a normal panel + CBS


Usage:
  sesame cnv [options] <target.cg> <segments.tsv> <bins.tsv>
  sesame cnv --probes [options] <target.cg> <segments.tsv> <probes.tsv>

  Copy number: regress the target's per-probe total intensity on a
  panel of normals (OLS), take log2(target/fitted) per probe, bin
  along the genome (median per bin), and segment by circular binary
  segmentation. The target may hold several samples -- one profile
  each. Writes the CBS <segments.tsv> plus a detail file that is
  per-bin by default, or per-probe with --probes; cinderplot reads the
  two as separate layers (segment step + points).

Arguments (all required):
  <target.cg>       total-intensity .cg, from `preprocess --raw-signal
                    --output total_intensity`; may hold several samples
  <segments.tsv>    output: sample chrom start end nbin seg.mean
  <bins.tsv>        output: sample chrom start end nprobes log2ratio
                    (--probes: sample Probe_ID chrom pos log2ratio)

Options:
  --probes          write per-probe rows to the detail file, not per-bin
  --exclude LIST    comma list of chromosomes to drop from both outputs
                    (e.g. chrY for a female sample; unreliable on arrays)
  --normals FILE    normal panel .cg (default: the store's
                    <P>.cnvnormals.cg, shipped with the annotation v8.2+)
  --platform P      EPIC | EPICv2 | HM450 | MSA; supplies the
                    defaults. A path to an ordering file also works,
                    for a custom array.
  --index FILE      ordering .tsv.gz (overrides --platform)
  --coords FILE     probe coordinate .tsv.gz (default: the store's)
  --genome BUILD    genome build (default hg38)
  --tilewidth BP    genome bin width (default 50000)
  --min-probes N    drop bins with < N probes (default 20)

Notes:
  coords, genome and the normal panel all default to the store for
  --platform; `yame fetch` brings the panel with the annotation.

sesame vcf — genotype the SNP probes into a VCF (formatVCF)


Usage:
  sesame vcf <prefix> [options]

  Genotype the SNP probes into a VCF (sesame formatVCF). Runs on the
  raw signal -- channel inference would erase the Type-I genotype.
  <prefix> is an IDAT pair. Sites-only VCF on stdout.

Options:
  --platform P      EPIC | EPICv2 | HM450 | MSA (else from the bead
                    count), or the path to an ordering file, which is
                    how a custom array is named.
  --index FILE      ordering .tsv.gz (overrides --platform)
  --snp FILE        SNP table (default: the store's
                    <P>.<genome>.snp.tsv.gz)
  --genome BUILD    genome build (default hg38)
  --variants        keep only rs and channel-switching probes, dropping
                    the non-switching Infinium-I bulk (unlikely to vary)
  --min-beads N     mask probes with < N beads (default 0)
  --out FILE, -o    write here instead of stdout

sesame region — a region's betas as long-form TSV, for plotting


Usage:
  sesame region <chr:beg-end> --betas <beta.cg> [options]
  sesame region --gene <NAME> --betas <beta.cg> [options]

  Extract a region's betas from a multi-sample beta.cg as long-form
  TSV: chrom beg end Probe_ID beta sample_name -- one row per (probe
  in region) x sample, the plot-ready feed for a locus view. Give an
  explicit chr:beg-end (commas ok) or a gene symbol via --gene, whose
  span is looked up in the GENCODE gene models.

Options:
  --betas FILE      multi-sample beta.cg (required)
  --gene NAME       resolve the region from a gene symbol (e.g. ADA)
                    rather than a chr:beg-end
  --gene-models F   GENCODE genes.bed.gz for --gene (default: store)
  --pad BP          extend the region by BP on each side (default 0)
  --platform P      EPIC | EPICv2 | HM450 | MSA, or the path to an
                    ordering file. Supplies Probe_IDs and coords.
  --index FILE      ordering .tsv.gz (overrides --platform)
  --coords FILE     probe coordinate .tsv.gz (default: the store's)
  --genome BUILD    genome build (default hg38)
  --out FILE, -o    write here instead of stdout

sesame mliftover — lift a .cg/.cm across arrays, or to/from a genome (hg38)


Usage:
  sesame mliftover --to <platform|genome> [options] <in.cx> <out.cx>

  Lift a .cg between row spaces (mLiftOver). Array to array joins on
  probe-ID prefix: EPICv2/MSA carry a _suffix that EPIC/HM450/HM27
  lack, stripped on the modern side when crossing families; each
  target probe takes the first prefix-matched source beta (NA if
  none). Array to GENOME (--to hg38) joins on coordinate, into the
  genome's CpG universe (one row per CpG). Only cg probes cross, by
  ID -- rs/nv carry genotype, not methylation, wherever they map;
  replicates on one CpG are averaged; cg probes with no CpG row are
  dropped. Every count is reported.
  Genome to array (--platform hg38 --to MSA) takes any genome-indexed
  .cg or .cm mask down to one row per probe; that direction is
  lossy (CpGs with no probe are gone). Output is always positional
  to the TARGET ordering.

Options:
  --to P            target: EPIC | EPICv2 | HM450 | MSA, or a genome
                    such as hg38 whose cpg_nocontig.cr is in the store (required)
  --platform P      source platform or genome, setting the direction
                    (inferred from --index's filename if omitted)
  --simulated-depth N write betas as format 3 with M+U=N on every covered
                    row and 0,0 elsewhere -- what a whole-genome model
                    (methscope classify) reads. Default: keep the input format.
  --coords FILE     the array side's <plat>.<genome>.coord.tsv.gz
                    (default: the store's)
  --threads N, -t   worker threads sharing the one map in-process
                    (default: number of CPUs); output is identical at any N
  --index FILE      source ordering .tsv.gz (default: the store's)
  --index-to FILE   target ordering .tsv.gz (default: the store's)

Notes:
  The empirical liftOver quality mappings and impute=TRUE are not
  ported.

sesame impute — fill missing betas (matrix mean or genomic neighbours)


Usage:
  sesame impute --method <mean|neighbors> [options] <in.cg> <out.cg>

  Fill missing (NA) beta values in a beta.cg (imputeBetas*). Output is
  positional to the same ordering.

--method mean:
  imputeBetasMatrixByMean -- replace each NA with a mean.
  --axis probe      mean of the probe across samples (default; R axis=1)
  --axis sample     mean of the sample across probes (R axis=2)

--method neighbors:
  imputeBetasByGenomicNeighbors -- replace each missing probe with the
  mean of its nearest non-missing genomic neighbours.
  --platform P      EPIC | EPICv2 | HM450 | MSA (for the coord table)
  --coords FILE     probe coordinate .tsv.gz (default: the store's).
                    Matches R for mapQ>=1 probes; differs only on
                    mapQ=0 (multi-mapping) ones.
  --genome BUILD    genome build (default hg38)
  --max-neighbors N up to N nearest neighbours (default 3)
  --max-dist BP     search window in bp (default 10000)

Inspect / convert

sesame describe-probe — probe IDs for a positional file, or their coordinates


Usage:
  sesame describe-probe [options] <file>

  Say what a positional file's rows are, or where probes are.

  Default: prepend the ordering's Probe_ID to a positional file's rows,
  as TSV on stdout. <file> is a YAME .cg/.cm/.cx (fmt0 mask, fmt3 M/U
  or --beta, fmt4 float) or a text .tsv[.gz], e.g. a .hg38.coord.tsv.gz.
  The row count must match the ordering -- same platform and tag that
  made the file.

  With --genome: <file> is instead a list of probe IDs (one per line,
  "-" for stdin) and the output is Probe_ID <TAB> <chrm>_<beg1>, in
  input order, no header. That second column is what `yame rowsub -L`
  reads, so the genomic neighbourhood of a probe set is:

    sesame describe-probe --platform HM450 --genome hg38 probes.txt \
      | cut -f2 | yame rowsub -L - -w 15 -M map.tsv -1 wgbs.cg > win.cg

  A bare cg number matches every design suffix (EPICv2/MSA), so one
  query can yield several lines. An ID the platform does not carry is
  an error. Probes with no CpG in the genome, or on a contig a
  genome-indexed store cannot address, are dropped and counted on
  stderr.

Options:
  --platform P      EPIC | EPICv2 | HM450 | MSA. Omit it and the
                    ordering comes from the filename, else from the
                    file's row count -- each ordering has a distinct
                    length, so a positional .cg identifies itself.
  --index FILE      ordering .tsv.gz (overrides --platform)
  --with LIST       also emit these ordering columns after the ID:
                    M, U, col, mask (or all). The ordering is already
                    open -- this saves joining it back on.
  --all, -a         emit every sample column (default: first only)
  --beta            read fmt3 M/U as beta
  --no-header       omit the header row
  --genome BUILD    coordinate mode: read probe IDs, print
                    <chrm>_<beg1> per probe (e.g. hg38)
  --coords FILE     the coordinate table (default: the store's
                    <platform>.<genome>.coord.tsv.gz)

sesame idat-dump — dump raw IDAT records, or a summary header


Usage:
  sesame idat-dump [options] <file.idat|file.idat.gz>

  Print IDAT contents. Default prints a summary header.

Options:
  --tsv             emit addr<TAB>mean<TAB>sd<TAB>nbeads
  --head N          limit to the first N records

sesame deidentify — remove the genetic fingerprint (SNP probes)


Usage:
  sesame deidentify [options] <prefix|idat> [<out.idat>]

  Strip the genetic fingerprint from an IDAT: the SNP (rs) probes'
  bead Mean intensities are zeroed (default), or reversibly scrambled
  with --randomize --seed. Everything else is copied byte-for-byte, so
  the file reads identically apart from the SNP beads (ports sesame's
  deIdentify). A <prefix> processes both Grn and Red; a single
  .idat[.gz] just that file. Output defaults to <name>_noid_*.idat,
  or _reid_ with -r.

Options:
  --randomize       scramble the SNP means instead of zeroing
                    (reversible)
  -r, --reidentify  reverse a --randomize; needs the same --seed
  --seed N          seed for --randomize / -r
  --platform P      EPIC | EPICv2 | HM450 | MSA (else from the bead
                    count), or the path to an ordering file, which is
                    how a custom array is named.
  --index FILE      ordering .tsv.gz (overrides --platform)
  --out FILE        output path (single input only)

Notes:
  --randomize uses this tool's own PRNG, so a scrambled file
  round-trips with `deidentify -r --seed <same N>`, not with R.
  Zeroing is R-equivalent and irreversible.

Other

sesame version — version, the yame it links, the store

sesame 2.2.0
    built against  YAME v1.57
    store          /home/user/.local/share/yame   ($YAME_DATA_HOME unset; yame fetch fills it)

Citation — the SeSAMe papers

SeSAMe2 is a C reimplementation of the SeSAMe method family. Please cite the work behind the parts you use:

Infinium annotation & masking — Zhou, Laird & Shen. Nucleic Acids Research 2017. 10.1093/nar/gkw967
SeSAMe / pOOBAH — Zhou, Triche, Laird & Shen. Nucleic Acids Research 2018. 10.1093/nar/gky691
mLiftOver — Chen & Zhou. Bioinformatics 2024. 10.1093/bioinformatics/btae423
KnowYourCG (KYCG) — Goldberg et al. Science Advances 2025. 10.1126/sciadv.adw3027
MSA array — Goldberg et al. Cell Genomics 2025. 10.1016/j.xgen.2025.100929
ELBAR / low-input — Lee et al. Nucleic Acids Research 2024. 10.1093/nar/gkae127
Cross-species analysis — Ding, Kaur, Horvath & Zhou. Briefings in Bioinformatics 2023. 10.1093/bib/bbac617
MM285 mouse array — Zhou et al. Cell Genomics 2022. 10.1016/j.xgen.2022.100144