Guide to NGS Platforms

The development of next-generation sequencing (NGS) over the past 15 to 20 years has seen multiple technologies rapidly leapfrogging each other’s milestones to sequence more DNA faster, cheaper, and more accurately. While Illumina sequencers still predominate, other NGS providers are gaining ground, providing important alternative options for sequencing mechanisms, read lengths, run times, ease-of-use, and scalability. This article aims to guide researchers through the main differences among today’s common NGS platforms.

Short-read sequencing

NGS sequencing platforms can be roughly divided into short- (usually <300 bases) and long-read (about 1 kb and above) sequencing (also known as single-molecule sequencing) according to the size of DNA fragments sequenced. Short-read sequencing used to be clearly higher in throughput and lower in cost than long-read sequencing, but today—as the latter has quickly evolved—these determinations depend more on the individual project and its sequencing needs.

Illumina

Illumina’s sequencing methods use a proprietary cluster generation process and sequencing-by-synthesis with fluorescently-tagged chain terminators. This has become the most widely adopted and published sequencing method, with more than 300,000 peer-reviewed publications. The longest read length of current Illumina platforms is 2x300, but the upcoming launch of Illumina^® Complete Long-Read technology will enable researchers to generate contiguous long reads of 6–7 kb consistently, and up to 30 kb, according to Brooke Murphy, Director of Product Marketing at Illumina. For current technology, run times can vary from a few hours to 2 days, depending on the research project, and the instrument used. “Many of our instruments can perform a sequencing run in about a day,” says Murphy. “Output in sequencing reads can range from millions to over 20 billion depending on the customer’s needs.”

With over 20,000 NGS systems shipped worldwide, Illumina sequencers are currently more widely adopted compared to NGS platforms. Thus, an advantage of using an Illumina platform is the numerous published protocols, and greater knowledge and support in research communities worldwide. “Our customers have access and engagement with the largest ecosystem of applications, protocols, library prep, analysis options, and technology users to propel genomic research,” says Murphy.

Thermo Fisher Scientific

Thermo Fisher Scientific’s Ion Torrent sequencing platforms perform NGS by measuring pH changes across millions of wells on a semiconductor chip during a sequencing run, with read outputs depending upon the Ion chip. Read lengths depend on the Ion Torrent system used: the Ion Torrent Genexus Integrated Sequencer is capable of up to 400 bp read lengths, and the Ion GeneStudio S5 System can produce up to 600 bp read lengths.

Run time on the Ion Torrent Genexus Integrated Sequencer varies from 14–24 hours (from nucleic acid input to report) and produces 15–60 million reads, depending on the assay type and sample batch size. Run times on the Ion GeneStudio S5 Systems vary from 3–24 hours and produce 2–130 million reads depending on the chip, according to Andrew Hutchison, Associate Director of Product Management, Clinical Sequencing Division at Thermo Fisher Scientific.

MGI/Complete Genomics

MGI, a global supplier of NGS and automation products, offers short-read sequencing products that have recently become more commercially available. They use DNA nanoball sequencing, in which fragments of gDNA or cDNA sample are amplified by rolling circle replication into DNA nanoballs, which are then arrayed at high density for sequencing.

MGI platforms have maximum read lengths of SE400 for single reads and PE300 for paired reads. Although the run time is longer, the high-density arrays yield many reads per run. “Typical run time of PE150 is within 3 days, and the newest model of G99 runs PE150 in 12 hours,” notes David Daly, West Coast District Sales Director at Complete Genomics (part of MGI). Depending on sequencer model, throughput is about 5 million to 360 billion reads per run.

Single-molecule sequencing

As so-called third-generation sequencing, single-molecule techniques are newer but gaining ground as researchers become more familiar with them, and as their costs decline. Besides sequencing longer DNA molecules, single-molecule or long-read sequencing offers other advantages, including the ability to distinguish large structural variations, regions of high homology or repetition, and splice variations. Single-molecule sequencing can also detect modifications such as DNA methylation during regular sequencing runs (short-read NGS requires special library prep for methylation detection), which is valuable for studying epigenetics. Disadvantages used to include lower accuracy, but recent and ongoing advances in this area are reducing error rates (for example, accuracy can now exceed 99.9%, rivaling that of short-read platforms).

PacBio

PacBio’s Single-Molecule Real-Time (SMRT) technology sequences single DNA molecules immobilized at the bottom of tiny microwells using fluorescently tagged bases. PacBio’s HiFi sequencing, which uses circular consensus sequencing for 99.9% accuracy, offers long reads even in difficult-to-sequence regions of the genome.

Read length “depends on library insert size, but is typically 15 kb, and typical maximum read length is approximately 40 kb,” says Aaron Wenger, Director of Product Marketing at PacBio. On the PacBio Sequel IIe system, run time is 30 hours, with the number of reads depending upon the library. “Read count varies by library type: 2 million for whole-genome sequencing, and 40 million for MAS-Seq RNA-sequencing libraries,” he says.

Oxford Nanopore Technologies

The sequencing technology of Oxford Nanopore Technologies centers around passing a single strand of DNA through a protein nanopore. Because the ionic conductivity of the pore changes depending on which nucleotides are within, the DNA sequence can be read by the ionic current moving through the pore. “Oxford Nanopore’s platform can sequence a DNA or RNA fragment of any length, from 30-base fragments, to full-length transcripts or millions-of-bases fragments for easy and complete assembly of telomere-to-telomere genomes,” says Rosemary Sinclair Dokos, Senior VP of Product Management at Oxford Nanopore Technologies. Because it can access all of the genome—sequencing as much as 14 Tb in one device run—the accuracy of Oxford Nanopore’s sequencing can exceed that of short-read platforms for capturing structural variants or copy number variations.

Oxford Nanopore’s platform doesn’t have a set run time; instead, “devices can be run for as little or as long as required to resolve the question or generate the required data,” says Sinclair Dokos. In addition to the ultra-high throughput formats, Oxford Nanopore’s platform offers a portable, handheld instrument to support sequencing in a wider range of applications. “Devices can be used outside the traditional lab environment—taking the analysis to the sample,” she says.

Additional considerations

In general, NGS providers continue to work on reducing costs, and increasing accessibility and scalability, with the goal of making NGS easier to use for more labs and applications.

Costs

Having been available longer, short-read sequencing is more prevalent and often less expensive (though single-molecule sequencing costs are becoming more competitive). Continuing the trend of the past few years, the costs of NGS continue to decline. “Illumina is committed to reducing the cost of sequencing, and has lowered the cost of sequencing by more than 500 times since our first platform, when comparing NovaSeq 6000 versus our first platform,” says Murphy. “The price per Gb on NovaSeq 6000 is less than $5/Gb., and on the newly launched NovaSeq X less than $2/Gb, which represents a 1500x reduction vs. our first platform.” Regardless of the sequencing platform used, costs greatly depend upon individual projects, the number of reads per sample, and NGS usage.

Ease-of-use and scalability

In contrast to earlier sequencers designed to maximize throughput, some newer models emphasize ease-of-use and scalability. For example, to support the use of in-house NGS by smaller labs and hospitals, Thermo Fisher offers the Genexus System to automate the entire NGS workflow, from sample extraction to report, with only 20 minutes of hands-on time. The Genexus System consists of the Genexus Purification System for extraction and quantification of nucleic acids, and the Genexus Integrated Sequencer, which automates library preparation, sequencing, and analysis within a single day. “The Genexus System offers a simple, fast, scalable solution for new-to-NGS labs to bring NGS in-house cost effectively,” says Hutchison.