Protocol Preview: Long-read sequencing, Nature’s Method of the Year

Blog post by Jasmin Skinner

While the 2022 calendar year was filled with extensive scientific discovery and innovation, one method in particular made a substantial impact within the life sciences community. Long-read sequencing (LRS), also known as third-generation sequencing, has been named 2022’s Method of the Year by Nature Methods, and with very good reason (1). In this blog post, we’ll discuss what exactly long-read sequencing is, how it was developed, and some of its potential research uses and applications.

Long-read sequencing is a recently developed method of genomic analysis, capable of sequencing strands of DNA that are 10,000-100,000 nucleotides in length (2) . Unlike short-read sequencing, which requires the genome to first be broken into small fragments of less than 1000 base pairs, LRS is capable of producing long, continuous strands of DNA without the need for amplification (2). This has allowed researchers to sequence areas of the genome that have been previously ‘dark’, such as the highly repetitive sequences found in telomeres (3). Previous technologies were unable to analyze these portions of the genome, making it near-impossible to study certain complex structures.

Next-generation sequencing (NGS), the predecessor of long-read sequencing, represented a significant methodological achievement when it was first introduced 20 years ago (1). It allowed scientists to accomplish in a single day what had previously required more than a decade of manual effort (4). However, despite the numerous benefits, it does have a number of limitations. Due to the sheer size and complexity of many genomes, it was difficult to generate complete genomic sequences with NGS, often resulting in many missing parts or errors (1). As the study of genomics fundamentally relies on identifying variants/mutations in the genome, any error introduced during sequencing could invalidate the results. This motivated the development of a sequencing technology that was capable of creating long, accurate sequences of DNA while also being able to detect some of the genomic complexity that was beyond the capabilities of previous methods (1).

The two long-read sequencing technologies that currently dominate the market are both capable of traversing the more repetitive regions of the genome, but their differing methods of sequence detection impact their read length, accuracy, and throughput (2). Pacific Biosciences Single Molecule Real-Time (SMRT) sequencing incorporates fluorescently labeled dNTPs into the nascent strand, after which point they are detected following laser excitation (2). The fluorophore is then cleaved from the nucleotide before the next dNTP is added to the strand so that only a single base pair is recorded at a time. This technique is capable of sequencing a DNA insert of anywhere between 1 kb to over 100 kb, greatly surpassing the less than 300 base pair read derived with a previous short-read sequencing technology,  Illumina (2).

The other major sequencing technology is produced by Oxford Nanopore Technologies (ONT), and utilizes a rather innovative and unique methodology. Linear DNA molecules are first attached to an adapter preloaded with a motor protein, and then loaded into a flow cell containing potentially thousands of nanopores embedded within a synthetic membrane (2). The motor protein unwinds the DNA and, coupled with an electric current, drives the DNA through the pore at a controlled rate. Depending on the nucleotide that passes through the pore a specific disruption in the current occurs, which allows the sequence of the DNA to be determined in real time. ONT sequencing has been capable of producing reads of over 1 Mb in length, over ten times longer than what is capable with Pacific Bioscience’s SMRT sequencing (2).

While these two techniques both produce longer strands than were previously possible, their differences in sequencing principles results in yield reads with varied lengths, error rates, and throughputs (1). Neither technology is without its respite drawbacks, but one technique may suit a particular research project better than the other.