Oxford Nanopore sequencing
Learning Objectives
- Understand the electrophysiology of protein nanopores (R10.4.1) and translocase motor stepping rates.
- Decode raw ionic current squiggles stored in POD5 files into nucleotide sequences using Dorado neural network basecallers.
- Perform direct native DNA and RNA base modification detection (5mC, 5hmC, 6mA, m6A) without bisulfite damage.
- Implement real-time targeted adaptive sampling ("Read Until") on MinION and PromethION platforms.
1. Physical Principles of Resistive Pulse Nanopore Sensing
Oxford Nanopore Technologies (ONT) operates on a completely non-optical physical principle: direct electrical resistive pulse measurement.
An electrically resistant synthetic polymer membrane separates two electrolyte chambers (cis and trans). Engineered biological protein nanopores (such as the modern R10.4.1 pore featuring dual constriction reader heads) are embedded within the membrane. When a constant voltage potential (typically -180 mV) is applied across the membrane, ions flow through the pore aperture, generating a baseline ionic current of approximately 100 to 200 picoamperes ($ ext{pA}$).
A molecular motor protein (a processive helicase) bound to a sequencing adapter ratchets single-stranded DNA through the pore lumen at a steady rate of approximately 400 bases per second. As nucleotides occupy the pore constriction, they physically obstruct the ionic aperture, disrupting the flow of ions. Because each combination of k-mers presents a distinct geometric and electrostatic profile, it produces a characteristic current deflection (a "squiggle").
2. Dorado Basecalling & Direct Native Epigenetics
Unlike optical sequencing platforms where native chemical modifications are erased during PCR amplification, nanopore sequencers directly sense intact, unamplified native genomic DNA and native mRNA molecules:
# 1. High-accuracy GPU basecalling with modified base calling
dorado basecaller sup,5mCG_5hmCG \
pod5_data/ \
--device "cuda:0" > calls.bam
# 2. Align long reads to human reference genome
minimap2 -ax map-ont -t 16 GRCh38.fa calls.bam | samtools sort -@ 8 -o aligned.bam -
samtools index aligned.bam
# 3. Aggregate 5mC and 5hmC methylation frequencies with modkit
modkit pileup aligned.bam methylation.bed --ref GRCh38.fa --threads 16