assay · intermediate · partly written
Enzymatic methyl-seq (EM-seq)
EM-seq reads DNA methylation as a C-versus-T letter change like bisulfite, but with enzymes that first protect methylated cytosine and then deaminate the rest, instead of harsh chemistry.
Also called: EM-seq, enzymatic methyl sequencing
How it works
EM-seq protects first, then converts. One set of enzymes (TET2 and T4-BGT) shields methylated and hydroxymethylated cytosine. A second enzyme, APOBEC3A, then turns every unshielded cytosine into uracil, read as T. The result looks like a bisulfite run, without the harsh chemistry that destroys DNA. 1
Reported, not yet replicated
Independent head-to-head more
De Abreu and colleagues compared WGBS, the Illumina EPIC array, EM-seq and Oxford Nanopore on three human samples. EM-seq showed the highest concordance with WGBS and delivered consistent and uniform coverage; the authors attribute the concordance to similar sequencing chemistry. Each method found some unique CpGs. Concordance with WGBS is not proof of accuracy, since shared chemistry could share an error. Only the abstract was read. 2
Reported, not yet replicated
Against bisulfite more
Bisulfite sequencing degrades DNA severely under its best-conversion conditions; EM-seq is designed to avoid that and works from much less DNA. In the vendor paper EM-seq covered more CpGs at low depth, while WGBS covered more at 10x. Neither method separates 5mC from 5hmC alone. 13
Reported, not yet replicated
Question answered
The same question as bisulfite sequencing: which cytosines were methylated, per CpG, at single-base resolution, as a fraction of reads. The aim of the method is to answer it without the DNA damage bisulfite causes. 1
Reported, not yet replicated
Readout physical
A count of reads per cytosine position that read C versus T, as for bisulfite. Two enzymatic steps make the difference. First, TET2 oxidises 5-methylcytosine (through 5hmC and 5fC to 5caC) and T4-BGT glucosylates 5hmC; both products resist the next enzyme. Second, APOBEC3A deaminates the unprotected, unmethylated cytosines to uracil, which PCR copies as T. Unmodified C reads T; 5mC and 5hmC read C. The output runs through bisulfite analysis pipelines. 1
Reported, not yet replicated
Not measured
It reports 5mC and 5hmC together, as bisulfite does; the authors state an add-on step is needed to separate them. It reads a population fraction per site. It inherits a conversion-style error: protection that fails leaves 5mC exposed to deamination, and deamination that fails leaves C where T is expected. 1
Reported, not yet replicated
Setting
Sample prep
Standard protocol input of 10 to 200 ng DNA; a low-input protocol goes down to 100 pg. No bisulfite step, so no bisulfite damage is reported. 1
Reported, not yet replicated
Conditions
From the vendor paper: TET2 oxidised at least 99 percent of 5mC (the Discussion gives 99.5 percent). Residual 5mC that reached APOBEC3A was 0.9 percent for EM-seq versus 2.7 percent for commercial bisulfite kits (a comparison the paper quotes from earlier work). An unmethylated lambda control read under 0.6 percent methylated. 1
Reported, not yet replicated
Throughput
Coverage rather than sample throughput: at 1x depth EM-seq covered about 54 million of 56 million possible CpGs, against about 45 million for WGBS at 50 and 200 ng and about 36 million at 10 ng. At 10x depth WGBS covered more CpGs (615,413 versus 179,991). Both comparisons come from the vendor paper. 1
Reported, not yet replicated
Still open here
- Nearly all performance figures come from one vendor-authored paper. Independent evidence read here is one abstract (de Abreu 2025). A full-text independent benchmark is needed.
- A review attributes weak spots to the EM-seq authors (fully methylated XP12 DNA detected below expected, high duplicate rates at 100 pg) that were not found in the 2021 paper; not confirmed.
- A small cell-free DNA study reported lower EM-seq replicate correlation (confounded) and was seen only as a snippet; not cited.
- Throughput, cost per sample and time per run are not sourced here.
- TAPS, a third approach where the methylated base is the one that changes, was read only through an abstract record and is not covered.
Not yet written: Where in cascade, Throughput class, Time per run, Cost, Engineerable variants, Use cases, Before after, Capital, Emerging, ML role, Data availability.
Linked from
Sources
- Vaisvila, R., Ponnaluri, V.K.C., Sun, Z., Langhorst, B.W., et al.. Enzymatic methyl sequencing detects DNA methylation at single-base resolution from picograms of DNA. Genome Research (2021). paper; opened 2026-11-17 all records citing this
- de Abreu, A.R., Ibrahim, J., Lemonidis, V., et al.. Comparison of current methods for genome-wide DNA methylation profiling. Epigenetics & Chromatin (2025). paper; opened 2026-11-17 all records citing this
- Grunau, C., Clark, S.J., Rosenthal, A.. Bisulfite genomic sequencing: systematic investigation of critical experimental parameters. Nucleic Acids Research (2001). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)
This record has been changed 1 time, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.
- Status
- partly written
- Last updated
- 2026-11-17
- Record
- content/assays/enzymatic-methyl-seq.json