Most researchers first encounter m6A detection through antibody-based methods such as MeRIP-seq. These methods work by enriching methylated RNA fragments with an m6A antibody, then sequencing the enriched fraction. They are well established, but the antibody introduces known limitations — cross-reactivity with m6Am, batch-to-batch variability, and a requirement for relatively high input amounts that can exclude low-yield samples.

DART-seq offers an alternative route. Instead of pulling down methylated RNA with an antibody, it marks m6A sites enzymatically and reads them out through standard RNA sequencing. The method was developed by Kate Meyer's laboratory at Duke University and published in Nature Methods in 2019. Since then, it has been joined by a family of antibody-free m6A methods — including GLORI-seq, which adds single-base quantification on top of antibody-free detection. This article explains what DART-seq does, where it excels, where it falls short, and when GLORI-seq becomes the stronger choice.

Figure 1: DART-seq uses an APOBEC1-YTH fusion protein to mark m6A sites by editing adjacent cytosines, producing a detectable C-to-U change in RNA-seq data.

What Is DART-seq?

DART-seq — short for Deamination Adjacent to RNA Modification Targets — is an antibody-free method for transcriptome-wide m6A detection. It identifies m6A sites not by enriching methylated RNA, but by creating a detectable sequence change at positions next to m6A residues.

The core idea is simple. A fusion protein consisting of the YTH domain (which binds m6A) and APOBEC1 (a cytidine deaminase) is introduced into cells or applied to isolated RNA. The YTH domain guides the fusion protein to m6A sites, and APOBEC1 converts the cytidine immediately adjacent to the m6A residue into uridine. This C-to-U edit is then read out as a C-to-T change in standard RNA-seq data.

Because the edit only occurs at positions where the fusion protein binds — and the fusion protein only binds where m6A is present — each C-to-U event marks an m6A site. The technique does not require an antibody, an enrichment step, or a specialized sequencing platform. Standard short-read RNA-seq is sufficient, and long-read platforms such as PacBio can also be used to examine m6A patterns on individual transcript isoforms.

How DART-seq Works — The APOBEC1-YTH Mechanism

The DART-seq workflow has four stages:

Expression or addition of the fusion protein. In cellular DART-seq, the APOBEC1-YTH construct is expressed in living cells. In the in vitro variant, purified APOBEC1-YTH protein is added directly to isolated RNA, accepting as little as 50 ng of total RNA.

Binding and deamination. The YTH domain recognizes the m6A modification and positions APOBEC1 at the adjacent cytidine. Because the m6A consensus sequence is RRACH — where the final position is an invariant cytidine — the target cytidine is reliably positioned next to the modification. APOBEC1 converts this C to U.

RNA sequencing. Standard RNA-seq library preparation and sequencing follow. No enrichment, no antibody incubation, and no additional enzymatic steps are needed beyond the deamination reaction.

Bioinformatic detection. C-to-U (read as C-to-T) mutations are identified from RNA-seq alignments and filtered against controls — typically an APOBEC1-YTH mutant with a disabled m6A-binding domain, or a METTL3-depleted sample in which m6A is globally reduced. Positions enriched for C-to-T edits mark m6A sites.

Two controls give the method its specificity. An APOBEC1-YTH mutant that cannot bind m6A produces no targeted editing, confirming that deamination depends on m6A recognition. METTL3 knockdown or knockout eliminates the majority of m6A in the transcriptome, and DART-seq signals drop accordingly. Together, these controls distinguish genuine m6A-dependent editing from background APOBEC1 activity.

Key Advantages of Antibody-Free m6A Detection

Ultra-low input. DART-seq works from as little as 10 ng of total RNA in the cellular format, and approximately 50 ng for in vitro DART-seq. Antibody-based methods such as MeRIP-seq typically require 150–500 ng or more, making DART-seq accessible for samples where RNA yield is limiting — early embryos, flow-sorted populations, and microdissected tissue.

No antibody cross-reactivity. The most widely used m6A antibodies also recognize m6Am — a distinct cap-adjacent modification with different biological roles. This cross-reactivity means antibody-based datasets contain a fraction of m6Am signal that cannot be separated from genuine m6A. The YTH domain used in DART-seq does not bind m6Am due to hydrogen-bonding requirements with the 2′-OH of m6A, avoiding this confound entirely.

Compatible with standard and long-read sequencing. DART-seq output is standard RNA-seq data with C-to-T mutations. No specialized library preparation, enrichment protocol, or sequencing chemistry is required. Long-read DART-seq on the PacBio platform, demonstrated by the Meyer laboratory, reveals how m6A sites cluster on individual transcript isoforms — information lost in short-read experiments.

Cost and time. Eliminating the immunoprecipitation step saves both reagent cost and hands-on time. The in vitro format, in particular, can be completed with standard molecular biology equipment and a routine RNA-seq workflow.

Where DART-seq Falls Short

Motif dependence. DART-seq relies on the cytidine immediately following the m6A residue in the RRACH consensus. m6A sites that do not conform to this motif — or in which the adjacent cytidine is not accessible — are invisible to the method. While RRACH is the dominant m6A motif, not every m6A site fits it.

Indirect detection. DART-seq does not detect m6A directly. It detects a C-to-U edit at the neighboring position. This introduces an inference step: a C-to-U edit is taken as evidence of m6A at the adjacent adenosine. While the controls are well designed, the signal is one step removed from the modification itself.

No quantification. DART-seq reports whether an m6A site is present but not what fraction of transcripts carries the modification. The C-to-U editing rate does not translate into m6A stoichiometry in a straightforward way, because editing efficiency depends on APOBEC1 activity, fusion protein expression level, and local sequence context in addition to m6A occupancy. A low editing rate could mean low m6A, or it could mean limited APOBEC1 access.

Off-target editing. APOBEC1 has intrinsic deaminase activity that is not fully suppressed by fusion to the YTH domain. Background C-to-U edits occur at positions unrelated to m6A, and while control experiments subtract most of this noise, the remaining background limits sensitivity at lowly modified sites.

DART-seq vs GLORI-seq — A Method Comparison

GLORI-seq (Glyoxal and nitrite-mediated deamination of unmethylated adenosine) takes a different route to the same destination: antibody-free, transcriptome-wide m6A detection. Instead of an enzyme fusion, GLORI-seq uses a chemical reaction — glyoxal and nitrite — to deaminate every unmethylated adenosine in the RNA to inosine, which is read as guanosine during sequencing. m6A residues, protected by the methyl group, remain as adenosine and are read as A.

The result is a direct, single-base readout of m6A status at every adenosine in the transcriptome. Unlike DART-seq, which reports an indirect C-to-U edit, GLORI-seq measures m6A at the modification site itself.

Figure 2: DART-seq marks m6A indirectly through adjacent C-to-U editing, while GLORI-seq reads m6A status directly at the modification site by protecting m6A from chemical deamination.

The essential trade-off is between simplicity and information content. DART-seq is the simpler experiment — express a fusion protein, run RNA-seq, call edits. But it delivers less information per site: presence without quantification. GLORI-seq is chemically more complex and sequencing-depth intensive, but it returns a quantitative measurement at every adenosine the sequencing covers. For a deeper discussion of how these methods fit into the broader m6A detection landscape, see our guide to choosing an m6A mapping method.

Choosing Between DART-seq and GLORI-seq

The right method depends on what question you are asking and what material you have available.

Choose DART-seq when:

• Your primary question is “where is m6A?” rather than “how much m6A?”
• RNA input is severely limited — 10–50 ng — and you cannot pool samples
• You need long-read data to examine m6A patterns on individual isoforms
• You are screening conditions, time points, or genotypes and need a fast, low-cost survey
• Your samples cannot be processed through a chemical deamination workflow

Choose GLORI-seq when:

• You need to compare modification levels between conditions — “is m6A higher in treated versus control?”
• Stoichiometry matters — you want to distinguish a site modified at 10% of transcripts from one modified at 80%
• You are studying m6A dynamics where quantitative changes, not just site gain/loss, are the biological signal
• You plan to model the relationship between m6A level and RNA stability, translation, or other quantitative readouts
• Your target journal or reviewer expectations demand single-base resolution

If you are uncertain which method fits your project, the GLORI-seq service page describes the full workflow from sample submission to data delivery. CD Genomics also offers MeRIP-seq for projects where antibody-based enrichment remains appropriate, and SELECT-m6A sequencing for targeted validation of specific m6A sites.

Figure 3: A decision framework for m6A method selection based on research question, input availability, and quantitative requirements.

From Detection to Quantification — When to Move Beyond DART-seq

DART-seq answered an important question when it was published in 2019: can m6A be detected without antibodies? The answer was yes, and the method opened the door to m6A studies on samples that were previously inaccessible.

Since then, the field has moved beyond detection to quantification. Knowing that an m6A site exists is useful for cataloging the epitranscriptome. Knowing what fraction of transcripts carries the modification — and how that fraction changes under experimental conditions — is what drives mechanistic insight. GLORI-seq represents this shift. It retains the antibody-free principle of DART-seq but replaces enzymatic marking with a chemical conversion that returns a direct, quantitative readout. For a step-by-step explanation of the chemistry and workflow behind GLORI-seq, see our guide to GLORI-seq principles and experimental steps.

The trajectory is familiar in genomics: methods evolve from detection to quantification, and the richer dataset eventually becomes the default. Bisulfite sequencing replaced methylation-sensitive restriction enzymes for DNA methylation, and single-cell RNA-seq supplemented bulk RNA-seq for transcriptomics. In epitranscriptomics, single-base quantitative methods such as GLORI-seq are on a similar path.

That does not mean DART-seq is obsolete. For projects where input is scarce, where long-read isoform information is needed, or where a rapid survey is the goal, DART-seq remains a practical choice. The decision is not about which method is newer — it is about which method answers your question with the resources you have. For projects where antibody-free detection and minimal input are the priority, RNA modification services at CD Genomics support assay selection and study design across the full range of m6A methods, including DART-seq-appropriate workflows. If your question requires quantification, GLORI-seq provides the single-base resolution and stoichiometric measurement that current mechanistic studies demand.

FAQ

1. What is the difference between DART-seq and antibody-based m6A methods like MeRIP-seq?

DART-seq uses an APOBEC1-YTH fusion protein to enzymatically mark m6A sites, eliminating the need for an m6A antibody. This avoids antibody cross-reactivity with m6Am and works with as little as 10 ng of RNA, whereas MeRIP-seq typically requires 150 ng or more and enriches methylated RNA through antibody pull-down.

2. Does DART-seq provide quantitative m6A measurements?

No. DART-seq detects the presence of m6A — “is this site modified?” — but does not measure the modification level at each site. The C-to-U editing rate depends on multiple variables beyond m6A occupancy and cannot be converted into a reliable stoichiometric measurement. For quantitative, single-base m6A stoichiometry, GLORI-seq is the appropriate method.

3. Can DART-seq detect all m6A sites in the transcriptome?

No. DART-seq requires a cytidine immediately 3′ of the m6A residue for APOBEC1 to act on. This cytidine is present in the canonical RRACH m6A motif, but m6A sites that deviate from this consensus — or where the adjacent cytidine is structurally inaccessible — will not produce a detectable edit. GLORI-seq, by contrast, has no sequence restriction and can detect m6A at any adenosine.

4. Is DART-seq compatible with low-input samples?

Yes — this is one of DART-seq's main strengths. Cellular DART-seq works with as little as 10 ng of total RNA, and in vitro DART-seq accepts approximately 50 ng. This makes the method suitable for samples where RNA yield is limiting, such as sorted cell populations, microdissected tissue, or early developmental stages.

5. When should I switch from DART-seq to GLORI-seq?

Switch to GLORI-seq when your question shifts from detection to quantification — when you need to compare modification levels between conditions, measure stoichiometry, or track quantitative changes in m6A. DART-seq is appropriate when the goal is a qualitative survey of m6A site locations, especially from low-input samples or when long-read isoform-level information is needed.

References

1. Meyer, Kate D. "DART-seq: an antibody-free method for global m6A detection." Nature Methods, vol. 16, no. 12, 2019, pp. 1275–1280. DOI: 10.1038/s41592-019-0570-0
2. Liu, Cong, Hanxiao Sun, Yunpeng Yi, Weiguo Shen, Kai Li, Ye Xiao, Fei Li, et al. "Absolute quantification of single-base m6A methylation in the mammalian transcriptome using GLORI." Nature Biotechnology, vol. 41, 2023, pp. 355–366. DOI: 10.1038/s41587-022-01487-9
3. Shen, Weiguo, Hanxiao Sun, Cong Liu, et al. "GLORI for absolute quantification of transcriptome-wide m6A at single-base resolution." Nature Protocols, vol. 19, 2024, pp. 1252–1287. DOI: 10.1038/s41596-023-00937-1
4. Fei, Shengyi, Zheng William Fang, and Boxuan Simen Zhao. "Unraveling the RNA Tapestry: A Symphony of Innovations in m6A Research Technology." Israel Journal of Chemistry, vol. 64, no. 3–4, 2024. DOI: 10.1002/ijch.202400014

CD Genomics provides epitranscriptomics services for research use only and they are not intended for clinical diagnosis or treatment.