Nanopore DNA Sequencer
A single protein pore, no wider than a DNA strand, listens to the electrical whisper of each passing base — turning a molecule into a readable sentence.
A nanopore sequencer passes a single DNA (or RNA) strand through a protein channel set into a thin membrane. An applied voltage drives ionic current through the pore; as each nucleotide transits the narrow constriction, it partially blocks the flow in a characteristic way. A sensitive amplifier records the current trace in real time, and a basecalling algorithm decodes the sequence of blockades into A, T, G, C — or modified bases — directly, without chemical labelling or optical detection.
The workhorse protein is α-hemolysin (αHL) from Staphylococcus aureus — a heptameric mushroom-shaped channel that spontaneously inserts into a phospholipid bilayer. Its β-barrel lumen narrows to a ~1.4 nm constriction, just wide enough for single-stranded but not double-stranded DNA. A helicase motor protein bound to the DNA strand controls translocation speed, ratcheting the strand through the pore at ~450 bases per second. Newer pores — Oxford Nanopore's proprietary R10.4.1, derived from a different engineered protein — feature a longer, dual-constriction barrel that reads a shorter k-mer window, improving discrimination of difficult homopolymer sequences. The entire array sits on a disposable flow cell with 512 addressable sensor wells; the MinION instrument reads all 512 channels in parallel via USB.
In plain terms
Imagine a tiny hole in a wall just barely wide enough to push a strand of beads through, one bead at a time. Each bead shape blocks the hole a little differently, changing how much water trickles through. By measuring the water trickle, you can figure out what each bead looks like — and that is how you read the DNA code.
Also called: Oxford Nanopore MinION, nanopore sequencing, single-molecule strand sequencing.
Why it matters
- Reads DNA or RNA directly, base by base, with no amplification required — eliminating amplification bias and enabling detection of DNA methylation in the same pass.
- Read lengths are limited only by input fragment length; reads exceeding 4 megabases have been reported, enabling chromosome-scale assemblies from single molecules.
- The MinION weighs 87 g and is powered through a USB port, making it the first sequencer deployable in field hospitals, aircraft, and remote rainforests.
- Real-time basecalling during a run enables adaptive sampling — the sequencer can eject uninteresting molecules mid-read to enrich for targets, without any physical intervention.
- Direct RNA sequencing detects mRNA modifications (m6A, pseudouridine) that are invisible to short-read platforms that require cDNA synthesis.
Where it lives
- Oxford Nanopore MinION Mk1C — USB-powered, 512 channels, field-deployable (105 × 23 × 33 mm, 87 g)
- Oxford Nanopore PromethION 48 — 48 flow cells in parallel, used for population-scale human genome projects
- Oxford Nanopore GridION — 5 flow cells, rack-mounted clinical laboratory instrument
- Rapid outbreak sequencing platforms: SARS-CoV-2, Ebola, and mpox sequenced in field conditions with MinION
- Space-borne sequencing: MinION operated on the International Space Station (NASA, 2016) to demonstrate microbial identification in microgravity
Key numbers
- Pore constriction (αHL): ~1.4 nm (Narrow β-barrel zone; vestibule opens to ~3.0 nm)
- Translocation speed (R10.4.1): ~450 bases/s (Controlled by helicase motor; raw signal sampled at ~5 kHz)
- Raw read accuracy (R7, 2014): ~66–80% (Early access programme; 2D reads improved to ~85%; Mikheyev & Tin 2014, Mol. Ecol. Resour.)
- Raw read accuracy (R10.4.1 + V14, 2024): >99% (Q20+) (Single-read simplex; Kim et al., PLOS Biology 2024; ONT chemistry technical document)
- Raw read accuracy (Dorado v5 basecaller): 99.75% (Q26) (Oxford Nanopore Technologies accuracy page, 2025)
- MinION dimensions: 105 × 23 × 33 mm, 87 g (USB 3.0 powered, 5 W)
- Active sensor wells per flow cell: 512 (R10.4.1 MinION flow cell)
- Maximum read length reported: >4 Mb (Ultra-long read protocol, ONT documentation)
Surprising facts
- The α-hemolysin pore's constriction is 1.4 nm across — roughly five water molecules wide — and yet four DNA bases simultaneously occupy its sensing zone as each strand threads through.
- A MinION flow cell's 512 sensor wells, each reading a single DNA molecule at a time, can collectively sequence an entire bacterial genome to 100× coverage in under 30 minutes.
- Raw read accuracy climbed from roughly 66% read identity in June 2014 (R7 pore) to over 99% with R10.4.1 and V14 chemistry in 2024 — a dramatic turnaround driven by pore redesign and deep-learning basecallers.
- Oxford Nanopore's PromethION generates enough data to sequence a complete human genome (3 billion base pairs) to 30× coverage in a single day.
Open questions
- Systematic homopolymer errors that no simple engineering fix eliminates: When a DNA strand contains a run of identical bases (e.g., AAAAAAA), successive nucleotides produce nearly identical current levels — the pore effectively 'stalls' perceptually while physically advancing. Even R10.4.1's dual-constriction barrel, specifically designed to address this, leaves residual errors in homopolymers longer than about 10 bases. The errors are systematic rather than random, meaning consensus calling does not fully correct them at any depth of coverage. Why protein pore geometry does not produce distinguishable signals for extended homopolymers despite the different chemical contexts of each base remains an open problem in nanopore biophysics.
A short history
The idea of using a nanopore as a sequencing sensor was conceived independently by David Deamer (UC Santa Cruz) and George Church (Harvard) in the early 1990s. The pivotal experimental proof came in November 1996 when John Kasianowicz, Eric Brandin, Daniel Branton, and David Deamer published in PNAS that RNA and DNA homopolymers could be driven through a single α-hemolysin channel in a lipid bilayer, producing distinct ionic current blockades proportional to polymer length. Mark Akeson joined the collaboration after reading this paper, and the team spent the following decade engineering motor proteins to slow translocation to readable speeds. Oxford Nanopore Technologies, founded in 2005 as a spinout from Hagan Bayley's Oxford group, commercialised the approach. MinION was launched in an early access programme (MAP) in mid-2014; the first published genome assembled from MinION reads appeared in September 2014 (Quick et al., GigaScience). Successive pore generations — R7 (2014), R9 (2016), R9.4.1 (2017), R10 (2019), R10.4.1 (2022) — progressively improved accuracy, culminating in the Q20+ chemistry released with R10.4.1 and the V14 kit.
Sources
- Kasianowicz, Brandin, Branton & Deamer — Characterization of individual polynucleotide molecules using a membrane channel, PNAS (1996) (https://www.pnas.org/doi/10.1073/pnas.93.24.13770)
- Oxford Nanopore Technologies — sequencing accuracy page (R10.4.1 / V14 chemistry, 2025) (https://nanoporetech.com/platform/accuracy)
- Oxford Nanopore Technologies — Chemistry Technical Document (Kit V14 / R10.4.1) (https://nanoporetech.com/document/chemistry-technical-document)
- Mikheyev & Tin — A first look at the Oxford Nanopore MinION sequencer, Mol. Ecol. Resour. (2014) (https://onlinelibrary.wiley.com/doi/10.1111/1755-0998.12324)
- Loman, Quick & Simpson — A complete bacterial genome assembled de novo using only nanopore sequencing data, Nature Methods (2015) (https://www.nature.com/articles/nmeth.3444)
- Rang et al. — Oxford Nanopore R10.4 long-read sequencing enables near-perfect microbial genomes, Nature Methods (2022) (https://pmc.ncbi.nlm.nih.gov/articles/PMC9262707/)