Adaptive sampling – Oxford Nanopore

Adaptive sampling is an Oxford Nanopore Technologies (ONT) method for real-time targeted sequencing. The method enables enrichment or depletion of specific genomic regions during sequencing without additional library preparation steps.

Adaptive sampling uses a reference genome (FASTA) and user-defined regions of interest (ROIs, BED file) to selectively sequence DNA molecules in real time. Molecules that map to the target regions are retained, while off-target molecules are rejected from the pore. It can be used either in enrichment mode (selecting for specified sequences) or depletion mode (rejecting specified sequences)

Typical enrichment levels are in the range of 4–6x, depending on sample properties and target design.

Adaptive sampling can for example be used for:

• Targeted sequencing of defined genomic regions
• Enrichment of low-abundance sequences
• Analysis of structurally complex loci
• Depletion of unwanted sequences (e.g. host DNA in metagenomics)

Method description

During sequencing, the initial part of each DNA molecule is basecalled and aligned against the reference genome in real time.

• If the molecule matches a region of interest, sequencing continues
• If not, the voltage across the nanopore is reversed and the molecule is ejected

This process allows selective sequencing without additional wet-lab enrichment.

Both enrichment mode (retain targets) and depletion mode (reject specified regions) are supported.

Target design

Regions of interest (ROI)

The size and composition of the target regions strongly influence performance.

• Minimum target size: ~0.1% of the genome
• Recommended: ≤5% of the genome (including buffers)
• Upper limit: ~10%

Buffer regions

Flanking regions (“buffers”) are recommended to improve enrichment efficiency.

• Standard recommendation: ~10 kb upstream and downstream of each ROI
• Buffers: this increase the probability that a fragment is classified as on-target early during sequencing
• Use https://adaptive-sampler-check.streamlit.app/Documentation to check your BED files

Smaller target regions generally result in higher enrichment.

Buffers may also be used to ensure that the total targeted region reaches the minimum recommended size.

Sample requirements

Please see library preparation page, e.g. Nanopore DNA Sequencing.

• Fragment length should ideally be comparable to the ROI plus buffer size
• Slightly shorter fragments than typical long-read workflows may improve selection efficiency

Data output and performance

Adaptive sampling provides moderate enrichment of target regions.

• Typical enrichment: ~4–6x
• ONT-reported range: up to ~5–10x

Performance depends on:

• Target size and distribution
• Fragment length
• Genomic context
• Sample quality

Limitations

Adaptive sampling introduces a trade-off between enrichment and sequencing yield.

• Off-target reads are actively rejected
• Repeated rejection leads to faster pore degradation
• Total sequencing output is therefore reduced compared to standard runs

Careful target design is important to balance enrichment and yield.

Input files

The following files are required:

• Reference genome (FASTA format)
• Regions of interest (BED format)

Support for ROI and buffer design can be provided if needed.

Further information

• NGI Tech Note: Adaptive Sampling – targeted Oxford Nanopore long-read sequencing
• Oxford Nanopore Technologies documentation: Adaptive sampling
• https://adaptive-sampler-check.streamlit.app/ 

Last Updated: 18th June 2026