Nanopore sequencing
Nanopore sequencing
Nanopore sequencing is a single-molecule sequencing approach in which DNA or RNA molecules pass through nanopores and produce signal changes that are computationally converted into sequence information. In practical systems, a motor protein controls molecular movement, many pores operate in parallel, and neural-network basecallers translate raw signal into reads.
The scientific value of nanopore sequencing comes from four connected properties: long reads, real-time acquisition, direct electrical sensing, and portable instrumentation. These properties make it useful for genome assembly, structural-variant analysis, microbial surveillance, outbreak response, environmental sequencing, transcriptomics, and direct detection of some nucleotide modifications.
Principle
A nanopore sequencing measurement has three coupled layers:
- Molecular layer - a DNA or RNA strand is prepared with adapters and is guided through a pore by an enzyme or motor protein.
- Physical layer - bases or short k-mer contexts inside the sensing region modulate ionic current or another measurable signal.
- Computational layer - signal processing and basecalling convert time-series data into reads, quality scores, and downstream analysis files.
A single current level is usually not assigned to one isolated base. The measured signal depends on a local sequence context, pore chemistry, motor speed, temperature, buffer, electronics, and model version. This is why basecalling and modified-base detection are model-based tasks rather than simple lookup tables.
Workflow from sample to result
| Stage | Main question | Typical output | Quality checks |
|---|---|---|---|
| Sample and library preparation | Is the nucleic acid long, clean, and compatible with the kit? | Sequencing-ready library | Input mass, fragment length, inhibitors, adapter ligation efficiency |
| Sequencing run | Are enough active pores producing stable reads? | Raw signal and run metrics | Pore activity, read length distribution, yield, quality over time |
| Basecalling | What sequence best explains the raw signal? | FASTQ or BAM reads | Read accuracy, quality scores, model and chemistry compatibility |
| Alignment or classification | Where do reads originate? | Mapped reads, taxa, or contigs | Mapping quality, coverage, contamination, reference bias |
| Interpretation | What biological or technical conclusion is supported? | Variants, assemblies, methylation calls, reports | Controls, replication, thresholds, orthogonal validation when needed |
Strengths
- Long reads can span repeats, structural variants, haplotypes, plasmids, and complex genomic regions.
- Real-time sequencing allows analysis during a run and supports rapid decision-making.
- Direct sensing can preserve information about some DNA or RNA modifications.
- Portable instruments make field, teaching, and low-infrastructure workflows possible.
- Flexible throughput allows small targeted runs and larger genome-scale projects.
Limitations and error sources
Nanopore sequencing should not be treated as automatically superior to short-read sequencing. Its suitability depends on the question, sample, required accuracy, cost, and analysis pipeline. Important limitations include:
- Raw read errors and systematic context-dependent errors.
- Sample-quality sensitivity, especially for long-read applications.
- Chemistry, pore, and model version dependence.
- Homopolymer and modified-base interpretation challenges in some contexts.
- Computational requirements for basecalling, alignment, assembly, and variant calling.
- Need for careful controls in clinical, regulatory, or high-consequence settings.
Applications
- Genome assembly - long reads help resolve repeats and structural complexity.
- Structural variation - long reads can span insertions, deletions, inversions, translocations, and repeat expansions.
- Microbial genomics - rapid identification, antimicrobial-resistance analysis, plasmid reconstruction, and outbreak tracking.
- Metagenomics and environmental DNA - real-time classification of mixed samples, with careful contamination control.
- Transcriptomics and direct RNA - isoform-level information and RNA modification signals in selected workflows.
- Epigenetics and base modifications - detection of methylation and other modifications when models and controls are appropriate.
- Adaptive sampling - real-time enrichment or depletion of targets by deciding whether to continue sequencing a molecule.
Data and file types
Raw and processed nanopore data may include POD5 or FAST5 signal files, FASTQ reads, BAM or CRAM alignments, sequencing-summary files, modified-base tags, and run reports. Good analysis practice records the instrument, flow cell, chemistry, kit, basecaller, model version, reference database, and command-line parameters.
Good practice
- Match the protocol to the biological question before optimizing yield.
- Use controls to separate library, pore, sample, and analysis effects.
- Report chemistry, basecaller, model version, and analysis parameters.
- Inspect both representative reads and aggregate statistics.
- Validate high-impact conclusions with independent evidence when possible.
Selected papers and resources
- Deamer D, Akeson M, Branton D. Three decades of nanopore sequencing. Nature Biotechnology. 2016.
- Jain M et al. Nanopore sequencing and assembly of a human genome with ultra-long reads. Nature Biotechnology. 2018.
- Wick RR, Judd LM, Holt KE. Performance of neural network basecalling tools for Oxford Nanopore sequencing. Genome Biology. 2019.
- Wang Y et al. Nanopore sequencing technology, bioinformatics and applications. Nature Biotechnology. 2021.
- Kovaka S et al. Targeted nanopore sequencing by real-time mapping of raw electrical signal with UNCALLED. Nature Biotechnology. 2021.
- Oxford Nanopore Technologies. Dorado basecaller and related tools.