eduKate Learning Manual · Science Route · EKS-SR-NANOPORE-20260905
Evidence reviewed: 5 September 2026. A route from native DNA in a nanopore to ionic-current signal, basecalling and sequence inference.
Follow one tiny change in electrical current—and discover why a DNA sequencer reads a moving molecule as a changing signal rather than recognising letters printed on a strand.
Wait, What? A nanopore sequencer does not read one DNA base at a time like a barcode scanner.
DNA passing through a nanopore changes the ionic current through that pore. But the sensing region is influenced by a short stretch of neighbouring nucleotides, not by one isolated base floating alone. The raw measurement is therefore a time-varying electrical trace—a “squiggle”. Software infers the sequence that most plausibly produced that trace.
That distinction is the whole route: molecule → physical signal → computational inference. Confusing the signal with the final letters hides both the power and the uncertainty of the method.
The big question—and the direct answer
How can one segment of ionic-current signal become a DNA sequence call? A DNA strand is moved through a nanometre-scale pore in a membrane while ions carry electrical current through the pore. Nucleotides occupying the sensing region alter that current in sequence-dependent ways. The instrument records current over time. A basecalling model then maps patterns in the raw signal to a likely nucleotide sequence.
The measured observable is current. The called letters are model-derived. A genomic interpretation—such as whether a variant changes a gene—is a further layer again and belongs to genomics or medicine, not to this route.
Why this route is worth learning
Nanopore sequencing links electricity, membranes, molecular motion, information and machine inference in one compact scientific journey. It is also a useful modern example of why “the instrument measured it” can mean several different things.
The National Human Genome Research Institute’s nanopore glossary, updated on 4 September 2026, describes DNA strands passing through tiny membrane pores and producing minute changes in electrical current that are converted into base information. This route slows that sentence down so each handoff stays visible.
What you will learn
- what the instrument directly measures;
- why a pore current depends on local sequence context;
- how time order becomes sequence order;
- why basecalling is an inference problem rather than a lookup table;
- how noise, speed, modifications and model training can affect the result.
1. Primary foundation: information can ride on a changing signal
A doorbell carries information with an electrical signal: current changes, and a system interprets the change as an event. Nanopore sequencing is vastly more sophisticated, but the foundational idea is similar. A physical process changes an electrical measurement, and the pattern of change carries information about what caused it.
DNA is not converted into electricity. Ions already carry current through the pore. The DNA changes how easily that ionic current passes through the sensing region.
2. Secondary mechanism: the pore is both a passage and a sensor
A nanopore is embedded in a membrane that separates two electrolyte-filled regions. An applied electrical potential drives ions through the pore. When a nucleic-acid strand occupies the constriction, it perturbs the ionic current.
The strand must move at a rate that allows useful signal to be recorded. Molecular machinery associated with the pore controls translocation in practical sequencing systems. This manual keeps that at the conceptual boundary; it is not an assembly or operating guide.
The key scientific point is that the current depends on the molecular configuration within the sensing region. Because several neighbouring nucleotides influence that region, the signal is contextual. “A gives 80 picoamps and C gives 70” is not an adequate model of real sequencing.
3. JC depth: a time series becomes a sequence hypothesis
The recorder samples current repeatedly as the molecule advances. The result is a time series containing signal levels, transitions, dwell-time variation and noise. Basecalling software must infer which nucleotide sequence is most consistent with the observed trace.
Modern basecallers use statistical and machine-learning models because neighbouring sequence context, movement speed, molecular modifications and instrument noise all shape the signal. A 2026 Bioinformatics paper describing CurrentView states the relationship plainly: nanopore sequencing measures ionic current as native DNA or RNA moves through a biological pore, and the resulting current changes are inferred into sequence by a basecaller.
The word inferred is not a weakness. It tells you exactly where the software enters the evidence chain.
4. Follow one nanopore current signal
Molecular arrival: a DNA strand reaches the pore. Translocation: a short stretch occupies the sensing region. Physical effect: the strand changes ion flow. Electrical observation: the instrument records a current value at a particular time.
Context: neighbouring current samples and the local sequence environment help define the signal pattern. Basecalling: a computational model assigns likely bases and their order. Read: successive calls become a sequence read with quality information. Alignment or assembly: downstream software may compare the read with a reference or combine reads into a larger sequence.
Our traveller stops at the sequence-read boundary. Gene function, diagnosis, ancestry, pathogen identification and clinical action are separate owners and require their own evidence and authority.
5. Why “one current level = one base” fails
Imagine trying to identify one piano key while several adjacent strings resonate together. The measured sound reflects a local configuration, not a perfectly isolated source. A nanopore sensing region likewise responds to several nucleotides at once.
This creates overlap: different local sequences can produce similar signals, while the same nominal sequence can vary because of noise, movement and molecular state. Basecalling therefore benefits from surrounding context rather than treating each sample independently.
A 2025 basecalling study in BMC Bioinformatics described raw current measurements as the input from which a neural-network model generates DNA sequence and highlighted sequencing noise and molecular variation as persistent challenges. That is a modern example of the general rule: the mapping from physical signal to symbolic sequence is probabilistic.
6. Modified bases make the signal richer—and the inference harder
Chemical modifications can change nanopore current without changing the underlying A, C, G or T identity. This is scientifically valuable because the raw signal may contain information beyond the canonical sequence. It is also a source of error if a model has not learned how that modification changes the signal.
Research published in Nature Methods in 2025 showed that aligning raw nanopore signal to DNA and RNA sequence can improve modification analysis. A 2026 Nucleic Acids Research study also reported that different RNA modifications alter current, dwell time or basecalling accuracy in different ways. The lesson is not that nanopores “see every modification”; it is that chemical state can influence the measured observable and must be modelled carefully.
How do we know? Observation versus inference
Observation: current samples through time, instrument timing and associated channel metadata. Derived signal features: normalised current, event boundaries or alignments used by particular analysis pipelines. Inference: base sequence, modification state, alignment to a reference and larger genomic interpretation.
A basecaller can be extremely accurate while remaining an inference engine. Confidence improves when reads agree, quality metrics are good and independent sequencing or known reference material supports the result.
Worked reasoning: one unexpected current shift
Suppose one region of a read produces a current pattern different from the model’s expectation. A weak conclusion says, “There is a mutation here.” Better reasoning asks whether the change could come from ordinary noise, an altered movement rate, a modified base, pore-to-pore variation, a basecalling error or a genuine sequence difference.
Next, inspect neighbouring signal, compare multiple reads and test whether the same sequence difference appears consistently. If an independent method supports the same base change, the mutation explanation becomes stronger. The signal is a clue; replication and comparison decide how far the claim can travel.
Failure modes and model limits
Noise failure: electrical and molecular variation can blur distinctions. Context failure: a model that treats bases too independently misses neighbouring effects. Training failure: unfamiliar sequence or modification patterns may be handled less well. Alignment failure: an erroneous read can be placed in the wrong genomic location. Interpretation failure: a correct sequence can still be given an incorrect biological meaning.
Sequencing technology also evolves quickly. Accuracy claims depend on pore chemistry, software version, sample type and analysis settings. A current manual should therefore avoid freezing one performance number as if it were timeless.
Misconceptions—and repairs
“The pore recognises letters.” No. It measures ion-current changes caused by molecular occupancy.
“Each base has one fixed current.” No. Local sequence context and molecular state matter.
“The raw squiggle is already a DNA sequence.” No. A basecaller converts signal into sequence hypotheses.
“A called variant explains health.” No. Variant interpretation and medical significance require separate genomic and clinical evidence.
Checkpoints, WHY questions and answer key
1. What quantity is directly measured? 2. Why is the current not a one-base lookup code? 3. What does a basecaller do? 4. Name two alternative explanations for an unexpected signal. 5. Why is a sequence read not yet a clinical conclusion?
Answers: 1. Ionic current through time. 2. Several nearby nucleotides and molecular conditions influence the sensing region. 3. It infers nucleotide sequence from signal patterns. 4. Noise, altered translocation, modification or model error, among others. 5. Biological and clinical meaning requires additional evidence and authorised interpretation.
Singapore and the wider world
Singapore’s research and biomedical ecosystem makes sequencing a natural modern connection, but the classroom value is broader. Nanopore science demonstrates how physics and computation can share one measurement chain: ions move, current changes, software interprets, and a biological question is asked only after those stages are kept straight.
Evidence and safety boundaries
This page is educational and does not provide sample-preparation recipes, wet-lab protocols, pathogen workflows, clinical interpretation or diagnostic advice. Laboratory molecular biology, sequencing operations, bioinformatics pipelines and medicine retain their specialist ownership.
KNOW → CONNECT → EXPLAIN → APPLY → CHECK
Know pore, ion, current, DNA strand and basecaller. Connect molecular occupancy to electrical signal. Explain why context matters. Apply an alternative-explanation test to a strange signal. Check whether a statement refers to raw measurement, called sequence or biological interpretation.
eduKateAI Direction Graph and where to go next
DNA strand → nanopore sensing region → altered ionic current → sampled time series → contextual signal pattern → basecalling model → sequence read → alignment or assembly → genomic interpretation. Membrane biophysics owns pore mechanism; sequencing engineering owns instrument operation; bioinformatics owns basecalling and downstream algorithms; medicine owns clinical meaning.
Return to Science World for the larger evidence map. Compare this route with signal-based manuals such as One Scintillation Photon, where a physical event also becomes an electrical signal before interpretation.
Authoritative sources and further reading
Current public definition: NHGRI, Nanopore DNA Sequencing, updated 4 September 2026. Signal-to-sequence work: SqueezeCall basecalling study (2025) and CurrentView, Bioinformatics (2026).
Signal alignment and modifications: Uncalled4, Nature Methods (2025) and systematic nanopore RNA-modification assessment, Nucleic Acids Research (2026). Evidence statements reviewed through 5 September 2026.
Teaching Guide for Parents, Tutors and Teachers
At Primary level, use a simple sound-recognition analogy: the microphone measures pressure variation, not words. At Secondary level, replace the microphone with an electrical sensor and separate measured current from interpreted symbol. At JC level, introduce time series, overlapping context and probabilistic inference.
Give learners a hypothetical trace with one unusual region and four possible causes: sequence change, chemical modification, noise and altered movement. Ask what repeated observations would discriminate among them. This turns “AI reads DNA” into a proper evidence question.
For the independent return, ask learners to write three sentences beginning “The instrument measured…”, “The basecaller inferred…” and “The biologist concluded…”. If those sentences remain distinct, the learner has understood the route.