Adaptive sampling: redefining targeted sequencing

Targeted sequencing — the selective sequencing of specific regions of interest rather than entire genomes — is a critical technique across numerous applications, including clinical research, pathogen surveillance, and species identification. Traditionally, the methods used to achieve targeted sequencing require enrichment or depletion via library preparation before the sample reaches the sequencer. These library preparation techniques, whilst effective, are often prone to PCR bias, complex, costly, and time consuming.

What if targeted sequencing could truly mean targeted sequencing — the enrichment of targets of interest, or depletion of off-target regions, during sequencing itself, with no need for special library prep steps? In this Nanopore Know-How blog, we introduce adaptive sampling: a method unique to Oxford Nanopore sequencing that enables just that.

Why use targeted sequencing?

If you only require data from a fraction of a genome, or one genome from a mixed microbial community, sequencing the entire sample may provide insufficient data for characterising these targets. Additionally, extra sequencing runs would increase costs, extend turnaround times, and lead to more complex data management.

This is where targeted sequencing shines. Instead, by focusing sequencing on targets of interest, you achieve high depth of coverage of targeted regions via a faster and more efficient workflow than whole-genome sequencing.

How is targeted sequencing typically performed?

Traditional targeted methods utilise library prep techniques to enrich specific regions before sequencing. A commonly used approach is targeted PCR, also known as amplicon sequencing, which can offer a simple and cost-effective workflow. However, the trade-offs are that PCR leaves behind regions that cannot be amplified, such as GC-rich and repetitive sequences, restricts analysis to fragment lengths that can be processed by the polymerase, introduces PCR bias, and erases epigenetic modifications. When combined with legacy short-read sequencing technology, read lengths are further reduced, hindering the analysis of large regions of interest such as structural variants (SVs).

Hybridisation capture is another method of targeted sequencing, typically used for the enrichment of wider target panels. This technique also relies on PCR, and requires a complex, lengthy workflow. Depletion techniques can also be performed to remove unwanted DNA, such as host material in microbial samples, but these again require additional sample prep steps.

Why is adaptive sampling different?

Adaptive sampling changes the game. With this innovative method developed by the Nanopore Community^1–4, you can perform target enrichment or depletion entirely during sequencing. The software-based method makes it possible to perform targeted sequencing without sacrificing long reads or native DNA. This means you can target sequencing to previously inaccessible regions, including complex SVs, long repeats, and GC-rich sequences — and detect native DNA modifications alongside genomic variants in a single sequencing run.

Adaptive sampling is a fast, simple workflow that requires no PCR, probe panels, or extra wet-lab steps. Plus, with no need for complex panel optimisation, you can easily update targets by editing your file of coordinates in a matter of minutes.

How does adaptive sampling work?

Your Oxford Nanopore sequencing device is controlled via the easy-to-use software, MinKNOW — and it is here that your sequencing run, including adaptive sampling, is set up. After specifying whether you wish to enrich for or deplete your specified target sequences, all you need to provide is a BED file containing the coordinates for these targets and a reference for alignment. After this, your targeted sequencing run is ready to start.

The targeting itself is made possible by the real-time nature of Oxford Nanopore sequencing and data analysis, which allows a DNA strand to be basecalled whilst it is still passing through a nanopore. During adaptive sampling, the sequence at the beginning of each DNA strand is compared against your list of targets. If the sequence matches a target of interest or is not a region to be depleted, the strand is ‘allowed’ by the software to continue sequencing. If it is not a target of interest or is listed for depletion, the DNA molecule is ejected from the nanopore so that it can begin sequencing another molecule.

This process significantly increases the depth of coverage for your targets of interest across a sequencing run. Our teams have observed around a five to ten-fold enrichment, enabling 20–40x mean depth of coverage of up to 10% of the human genome from a single MinION Flow Cell.

With this technology, you can target large, complex panels across a genome, such as disease-associated genes — you can even enrich for huge targets such as entire chromosomes² with ease. Furthermore, with ‘depletion’ mode, you can perform host depletion, such as of human DNA from microbial samples, or deplete abundant microorganisms from a metagenomic sample to capture rarer or understudied microbes of interest^2,3.

Adaptive sampling can be used with any Oxford Nanopore sequencer, on either MinION or PromethION Flow Cells. To maximise output, we also recommend washing the flow cell and loading fresh library using the Flow Cell Wash Kit.

What about library prep?

No special library preparation is required for adaptive sampling. Your whole DNA sample is prepared for Oxford Nanopore sequencing in the same way as for a whole-genome experiment. The PCR-free Ligation Sequencing Kit only takes around one hour to prep native DNA for sequencing, preserving long fragments and base modifications. You can also prepare up to four samples for multiplexed sequencing on a high-output PromethION Flow Cell using the Native Barcoding Kit.

The adaptive sampling workflow utilises more input library than a standard Oxford Nanopore workflow, ensuring that plenty of on-target DNA molecules are available for sequencing. Light fragmentation of high molecular-weight DNA helps to further improve data outputs and can also help to increase enrichment.

How are the Nanopore Community using adaptive sampling?

Scientists are harnessing the unique benefits of adaptive sampling to enhance their research across a wide range of applications. Researchers have shown future potential of the technique to enable intraoperative classification of central nervous system tumours⁵ and cancer-associated pathogenic variants across the complete MSK-IMPACT⁶ panel without the need for lengthy hybridisation capture⁷. Adaptive sampling has also been used to genotype all known neuropathogenic short tandem repeat expansions, plus pharmacogenomic targets, in a single sequencing run⁸, showing its potential to tackle the ‘diagnostic odyssey’ faced by many families. Going beyond the human genome, the method has been used to target DNA from the foodborne pathogen Staphylococcus aureus without the need for culturing⁹, and Plasmodium falciparum parasite DNA from blood research samples¹⁰.

With this unique technology, the Nanopore Community continues to push the boundaries of what is possible with targeted sequencing.

How will you use adaptive sampling?

Find out more about targeted sequencing with Oxford Nanopore

Want to get started? View adaptive sampling documentation