The appeal of nanopore sequencing rests on a simple physical principle. As a native RNA molecule threads through a protein nanopore embedded in an electrically insulated membrane, it disturbs the ionic current flowing through the pore. Because each combination of nucleotides, known as a k-mer, produces a characteristic current signature, the raw squiggle of electrical data can in principle be decoded into sequence. Crucially, because the molecule is read directly rather than converted to complementary DNA and amplified, chemical modifications such as N6-methyladenosine (m6A), 5-methylcytosine (5mC) and pseudouridine leave their own fingerprints in the current trace. This is a capability that short-read platforms like Illumina fundamentally lack, since their reverse transcription and polymerase chain reaction steps erase the very epitranscriptomic marks researchers want to find.

The review traces how the technology matured from the portable MinION released in 2015, through the high-throughput PromethION and the budget-friendly Flongle adapter, to successive chemistry generations from R9.4.1 to the R10 series and the newer RNA004 flow cells. Equally important has been the evolution of the computational pipeline. Base calling, the conversion of raw current into nucleotide letters, has progressed from Hidden Markov Models to LSTM-based recurrent neural networks in ONT’s Guppy, through convolutional architectures in Bonito, to Dorado, which the authors note has been reported to reach around 99.5 percent raw read accuracy. Alignment tools have followed a parallel trajectory, from GraphMap’s gapped seeding strategy through the now-ubiquitous Minimap2, which the study describes as three to four times faster than mainstream short-read mappers at comparable accuracy, to specialists such as Winnowmap2, which achieved false-negative and false-positive rates of just 1.89 percent each on repetitive reference sequences, far outperforming Minimap2’s 39.62 and 5.88 percent respectively in the cited evaluation.

Between base calling and modification detection sits the re-squiggling step, where raw signals are re-aligned to the called sequence so that per-base current features can be extracted. The review dissects Tombo’s five-stage implementation, spanning genome mapping, signal normalization, event detection, sequence-to-signal assignment and the resolution of skipped bases, and notes that most re-squiggling tools share this conceptual skeleton. Quality control software such as NanoPlot, pycoQC, MinIONQC and ToulligQC then guards the integrity of the data before any biological interpretation begins. Together, the authors argue, this pipeline has matured into a sophisticated, community-driven framework that has steadily improved accuracy, scalability and accessibility across a wide range of applications.

The heart of the paper is its survey of modification detection methods, which have evolved through three distinct technological waves. The first wave relied on classical statistics: Nanocompore, xPore and Yanocomp employed Bayesian generative models built on Gaussian mixture models, while DiffErr and DRUMMER used G-tests and ELIGOS applied Fisher’s exact tests to compare error profiles between modified and control samples. The second wave brought machine learning and deep learning: Nanom6A applied XGBoost for single-base m6A detection, MINES used random forests on four specific sequence contexts, EpiNano turned to support vector machines, and m6Anet deployed a neural network for transcriptome-wide m6A quantification. Pseudouridine specialists emerged in parallel, with Penguin combining several machine learning models to reach 93.38 percent accuracy on a HEK293 benchmark, and NanoMUD using bidirectional LSTMs to detect both pseudouridine and its therapeutically important cousin N1-methylpseudouridine.

The third and most recent wave is the most ambitious. Multi-modification detectors such as ModiDeC, which identifies m6A, pseudouridine, Gm, m1A and inosine in a single pass, and RNANO, which leverages attention-enhanced multi-instance learning to detect seven modification types, are pushing toward comprehensive epitranscriptomic profiling. Perhaps most transformative is the emergence of modification-aware base callers, exemplified by m6ABasecaller and ONT’s own Dorado, which now supports modified base calling for m6A and pseudouridine directly. These tools fold modification prediction into the base calling step itself, bypassing computationally expensive re-squiggling and post hoc statistics, and can operate without matched control samples or prior motif knowledge. The catch, the authors caution, is that such models demand training data with precise ground-truth annotations of modification status at individual nucleotides, a resource that remains scarce.

To test whether this proliferation of methods translates into reliable science, the team benchmarked five m6A detection tools, m6Anet, MINES, nanom6A, DRUMMER and DiffErr, against an NGS-derived reference database stratified into four confidence tiers, using the HEK293_WT and HMEC_WT datasets from GEO accession GSE132971. The discrepancies were dramatic. In HEK293_WT, m6Anet predicted 36,688 m6A sites and MINES 35,870, while DRUMMER reported only 36 and DiffErr just 3. Even among the three prolific tools, only 5,088 sites were commonly detected, and a large share of their predictions mapped to the low and very low confidence tiers of the reference: m6Anet, for instance, placed 15.5 percent of its calls in the low group and 22.6 percent in the very low group, with only 1.6 percent reaching high confidence. The pattern repeated in HMEC_WT, where the three leading tools shared just 2,806 sites.

The pseudouridine benchmark was even more sobering. NanoMUD predicted 82,581 pseudouridine sites in HEK293_WT, yet only 57 of them, about 0.1 percent, overlapped with the NGS-based reference, and NanoPsu managed just 31 overlaps. In HMEC_WT, Penguin predicted a staggering 644,248 sites but overlapped the reference at only 74 positions. Cross-tool concordance was similarly dismal, with NanoMUD and NanoPsu sharing only 396 sites in the first dataset and Penguin and NanoPsu agreeing on a single site in the second. The authors conclude that inconsistencies in pseudouridine identification are even more pronounced than for m6A, underscoring an urgent need for improved cross-tool consistency and methodological refinement.

Why does this matter beyond the technical community? RNA modifications are not decorative. m6A, the most prevalent internal modification in eukaryotic messenger RNA, regulates splicing, nuclear export, translation and mRNA stability, and its dysregulation has been implicated in heart failure, cancers and reproductive disorders. Pseudouridine enhances RNA stability and base stacking, and it is the key modification in chemically stabilized mRNA vaccines; NanoMUD has already been applied to quality control in vaccine design contexts. The review also contrasts nanopore approaches with alternatives: liquid chromatography-mass spectrometry offers accurate quantification but no sequence information, antibody-based methods lack single-nucleotide resolution, and newer NGS techniques such as GLORI and eTAM-seq achieve absolute m6A quantification but remain constrained by short reads and indirect chemistry. Nanopore’s direct, long-read, single-molecule approach remains uniquely positioned, but only if the computational interpretation of its signals can be trusted.

The authors are candid about the remaining obstacles. Nanopore error rates still exceed the 0.1 to 1 percent benchmark of Illumina short reads, and because some pipelines infer modifications from base calling errors, this intrinsic noise directly undermines detection. Direct RNA sequencing also demands substantial input, with the RNA004 protocol specifying 300 nanograms of poly(A) RNA or a microgram of total RNA, a barrier for clinical biopsies and single-cell work. Throughput remains modest at one to three gigabases per flow cell, and reliable modification detection requires coverage of at least 30-fold, while deep learning pipelines demand serious GPU infrastructure. Yet the trajectory is encouraging: multiplexed direct RNA protocols, real-time selective sequencing tools like SquiggleNet, signal-level mappers like Sigmap, and the integration of large language models and ensemble learning all point toward a future in which the epitranscriptome can be read as directly, and as reliably, as the genome itself. For now, the study’s message to researchers is clear: choose your tools carefully, validate against orthogonal methods, and treat any single algorithm’s modification calls with healthy skepticism.

Subject of Research: Benchmarking computational tools for detecting RNA modifications using Oxford Nanopore direct RNA sequencing

Article Title: Nanopore direct RNA sequencing for RNA modification analysis: workflow assessment and computational tool benchmarking

Article References: Wu, Z., Li, J., Xia, R., Dai, J., Su, J., Meng, J., & Zhang, Y. (2026). Nanopore direct RNA sequencing for RNA modification analysis: workflow assessment and computational tool benchmarking. Advanced Biotechnology, 4(1), Article 9. https://doi.org/10.1007/s44307-025-00093-5

Image Credits: AI Generated

DOI: 10.1007/s44307-025-00093-5

Keywords: nanopore sequencing, direct RNA sequencing, RNA modifications, m6A, pseudouridine, epitranscriptomics, base calling, machine learning, deep learning, benchmarking, Oxford Nanopore Technologies, computational tools

The clinical problem the team set out to address is well known in genomic medicine. Molecular diagnosis of rare disease currently plateaus at roughly fifty percent of patients, even after whole exome or whole genome sequencing. A substantial share of those unresolved cases involves variants that alter splicing, the process by which RNA transcripts are cut and re-joined to produce mature messenger RNA. Predicting whether a DNA change actually disrupts splicing in a patient’s tissues remains notoriously difficult, and variants of uncertain significance accumulate in reports without a practical way to test their effects directly.

Short-read RNA sequencing, which fragments transcripts into small pieces before reading them, often cannot show how exons connect across a whole transcript. Long-read platforms from Oxford Nanopore Technologies solve this by reading RNA-derived molecules end to end, revealing full isoform structures. Until now, however, long-read RNA approaches have typically demanded large control cohorts, complex bioinformatics and tissue samples that are hard to obtain, keeping them confined to specialist research laboratories rather than routine diagnostics.

RAPID, short for RNA Analysis Pipeline for Integrated Diagnostics, was designed from the outset to be diagnostically deployable. It is fully modular, covering the entire sample-to-answer journey, from primer design targeting genes of interest through nanopore sequencing to reproducible single-sample interpretation. Cases entered the workflow after exome or genome sequencing and multidisciplinary team review had narrowed the search to five or fewer candidate genes, meaning the sequencing effort could be focused and fast. Crucially, the method relies on accessible tissues such as blood and does not require large control datasets for interpretation.

The study applied the workflow to six probands with suspected monogenic neurometabolic disease, and the outcomes illustrate three distinct diagnostic scenarios. In two cases, targeted long-read RNA sequencing confirmed pathogenic splice disruption at the transcript level, converting uncertainty into a resolved molecular diagnosis. In one case, the RNA evidence argued against a candidate gene, prompting its exclusion and redirecting the diagnostic search. In the remaining three cases, transcript-level findings prioritised further DNA investigation, refining how the variants of uncertain significance should be weighed.

Technically, the workflow captured amplicons spanning the relevant exons and splice junctions of each candidate gene, generating reads long enough to assemble near-full-length isoform structures. The team demonstrated reproducibility by showing consistent isoform composition for a control gene across commercial blood RNA samples and an independent public long-read dataset, and they characterised the minimum read depth needed to detect transcripts at low fractional abundance. Quality control metrics, including read length and depth, were achieved using standard long-read sequencing infrastructure within a clinically relevant timeframe.

The authors also mapped their RNA results onto the established variant interpretation framework used by clinical scientists, in which aberrant splicing leading to frameshifts or premature stop codons, particularly in transcripts subject to nonsense-mediated decay, can support pathogenic classifications. By providing direct, mechanism-level evidence rather than in silico predictions, the workflow strengthens this evidence hierarchy and gives diagnostic laboratories a practical route to resolving variants that would otherwise remain reportable only as uncertain.

The significance for patients and families is considerable. A definitive molecular diagnosis can end a diagnostic odyssey that sometimes lasts decades, inform prognosis, guide surveillance and treatment, enable accurate genetic counselling, and open doors to targeted therapies and clinical trials. For a field in which half of patients still leave the sequencing process without answers, a rapid, cost-effective test that interrogates RNA directly addresses one of the largest remaining diagnostic gaps.

The researchers argue that RAPID shows long-read RNA sequencing can be implemented within existing diagnostic infrastructure, offering a scalable path to routine transcript-level assessment in clinical genomics. As nanopore sequencing becomes more widespread in NHS and hospital laboratories, workflows of this kind could shift splicing variant interpretation from probabilistic prediction to direct functional measurement. Supported by funding from the charity Sparks and conducted with appropriate ethical approvals and participant consent, the study points toward a future in which reading the transcript itself becomes a standard step in solving rare disease.

The biological importance of splicing in human disease provides useful context for why this approach matters. Introns interrupt most human genes, and their precise removal depends on short sequence signals at exon boundaries that are frequently disrupted by single-nucleotide changes lying well outside protein-coding regions. Because such variants often sit in positions that protein-prediction tools ignore, they can be classified as benign or left as uncertain even when they silently abolish a transcript. Deep intronic changes can also activate cryptic exon inclusion, an effect essentially invisible to standard exome analysis, which is one reason splice-altering variation has been described as a substantial hidden burden within the undiagnosed fraction of rare disease cohorts.

The interpretation framework applied in the study reflects a broader international movement toward using RNA evidence in clinical classification. Guidelines from professional bodies now allow aberrant splicing demonstrated at the transcript level to contribute to pathogenic classifications, provided the altered transcript is shown to escape nonsense-mediated decay or to produce a clearly deleterious product. This matters because a large proportion of loss-of-function variants predicted to trigger nonsense-mediated decay are already treated as pathogenic by default; demonstrating experimentally that a variant of uncertain significance produces the same consequence effectively moves it into that well-established category. Conversely, showing that a candidate variant leaves splicing intact can be equally decisive, as the gene-exclusion case in this cohort illustrates.

Tissue choice is a central consideration for any transcript-based diagnostic test. Genes are not expressed uniformly across the body, and the reference resources that underpin transcript interpretation, such as population-scale tissue atlases, show that many disease-relevant genes have their highest expression in tissues that cannot ethically or practically be sampled. Blood, however, is accessible, and the workflow’s reliance on peripheral blood RNA, collected in standard preservation tubes and stored frozen, means the logistics resemble those of routine clinical phlebotomy rather than specialist tissue procurement. The supplementary analyses showing that isoform composition of a control gene is consistent across commercial RNA references and public datasets speak to the reproducibility that regulators and accreditation bodies would expect of a deployable assay.

The depth requirements characterised by the team address a practical question that any diagnostic laboratory must answer before adopting such a test: how much sequencing is enough. Because a deleterious transcript may represent only a small fraction of all transcripts from a given gene, particularly when nonsense-mediated decay degrades the abnormal product rapidly, sensitivity at low fractional abundance is essential. Mapping the relationship between transcript rarity and required read depth gives laboratories a principled basis for setting sequencing targets and quality thresholds rather than relying on ad hoc criteria.

The study also sits within a distinctive UK genomic infrastructure. Several participating families had previously been sequenced through national genome programmes, and the ethical frameworks governing those programmes permit recontact of participants for follow-up sampling, a mechanism that allowed the researchers to obtain fresh RNA from already-investigated patients. This recontact pathway, coordinated through clinical interpretation partnerships, represents a model for how long-read RNA follow-up could be layered onto existing genomic medicine services without requiring patients to restart the consent and recruitment process from scratch.

From a health-systems perspective, the economics of the approach deserve emphasis. Whole genome sequencing has become affordable at scale, but the interpretation bottleneck, not the sequencing itself, now limits diagnosis. A targeted assay that sequences only a handful of genes consumes modest sequencing capacity on instruments that many hospital laboratories already operate or can readily access, and the modular design means primer sets can be redesigned quickly for each new case. The near-full-length isoform output also reduces interpretive ambiguity, because the analyst sees the complete exon connectivity of each transcript rather than inferring it from fragmented short reads.

The neurometabolic focus of the cohort is itself informative. Leukodystrophies and related neurodegenerative conditions of childhood often present with nonspecific imaging findings and progressive symptoms, and several of the genes implicated in these disorders are known to harbour splice-disrupting variants that confound standard pipelines. For families facing such progressive conditions, the speed of the workflow is not merely convenient; an earlier molecular answer can determine eligibility for emerging therapies, some of which are most effective when started before irreversible neurological damage accumulates.

Looking forward, the study suggests a tiered model of genomic diagnosis in which short-read sequencing remains the first-line discovery tool, while targeted long-read RNA analysis serves as a rapid functional triage step for the subset of cases with splice-relevant candidates. The authors’ demonstration that results can be obtained within a clinically relevant timeframe, using standard infrastructure and single-sample interpretation without large control cohorts, positions transcript-level evidence to become a routine component of multidisciplinary review rather than a research exception reserved for specially funded projects.

Subject of Research: A targeted long-read RNA sequencing workflow for functionally resolving splicing variants in rare disease diagnosis

Article Title: RAPID: a targeted long-read RNA workflow for functional resolution of splicing variants in rare disease

Article References: Montgomery, K.-A., Macpherson, H., Anderson, C., Wade, C., Gustavsson, E. K., Lynch, D. S., Wilson, L. C., Davison, J., Wakeling, E., Tuschl, K., Houlden, H., Clement, E., Mills, P. B., & Ryten, M. (2026). RAPID: a targeted long-read RNA workflow for functional resolution of splicing variants in rare disease. Genome Medicine. https://doi.org/10.1186/s13073-026-01754-3

Image Credits: AI Generated

DOI: 10.1186/s13073-026-01754-3

Keywords: long-read RNA sequencing, Oxford Nanopore Technologies, splicing variants, rare disease diagnostics, variants of uncertain significance, neurometabolic disease, transcriptomics, molecular diagnosis, clinical genomics, whole exome sequencing, RNA splicing, diagnostic workflow