There’s an previous saying within the area of expertise: “No one ever obtained fired for getting IBM” — a reference to the corporate’s once-ubiquitous computer systems. Change IBM with Illumina, a biotechnology firm in San Diego, California, and the identical could possibly be stated of DNA sequencing at present.
Keith Robison, a computational biologist at Ginkgo Bioworks in Boston, Massachusetts, who writes about sequencing applied sciences on a weblog known as Omics! Omics!, says that for many laboratories, Illumina “is the actually protected wager on the market”. Nevertheless, IBM’s days of pc market dominance are nicely up to now, and Illumina now faces a number of rivals who wish to problem — and maybe unseat — the present big of the sequencing market.
Researchers, naturally, are paying consideration. Pedro Oliveira heads the DNA-sequencing lab on the French Nationwide Sequencing Middle, also referred to as Genoscope, in Évry. The lab just lately partnered with a number of massive European analysis tasks, together with the European Reference Genome Atlas, which can herald an anticipated workload of 4 genomes per week. Certainly one of Genoscope’s priorities might be to extend its arsenal of Illumina devices — however that received’t be the restrict of its procuring checklist, and Oliveira has a broad vary of platforms to think about.
Some devices use complementary approaches that generate long-sequence reads spanning hundreds of nucleotides, in distinction to Illumina’s ‘brief reads’, that are sometimes within the 100- to 200-base vary. However the previous yr has additionally seen the launch of practically half a dozen competing short-read methods, every touting their very own benefits when it comes to high quality, effectivity and above all, price. “These are thrilling moments that we’re dwelling in,” says Oliveira, “as a result of that is the start of low-cost sequencing.” However the vary of decisions could be intimidating and complicated, given that the majority scientists are nonetheless ready to see the precise information, and to evaluate how nicely these platforms match their tasks.
A protected wager
Illumina entered the sequencing market with the acquisition of an organization known as Solexa in 2007. Solexa’s ‘sequencing by synthesis’ (SBS) expertise exploits the identical equipment that manufactures DNA in dwelling cells. A template DNA strand is learn by a DNA polymerase enzyme, which sequentially tacks on nucleotides that complement the template strand.
Every of the 4 DNA constructing blocks — A, T, G and C — is coupled to a selected fluorescent color and a ‘terminator’ chemical group that pauses additional DNA synthesis. Delicate optics determine the added nucleotide from the ensuing fluorescence, after which the tag and terminator are eliminated and the cycle repeats. The entire course of happens in a wafer-like ‘circulate cell’, through which huge numbers of DNA targets are imaged concurrently, producing tens of millions and even billions of brief reads per run.
Methodology of the yr: long-read sequencing
That strategy has been staggeringly profitable. By one estimate, greater than 90% of the world’s sequencing information as of 2022 had been generated on Illumina machines (see go.nature.com/3abj7ng). Dozens of would-be rivals have emerged to problem Illumina over time, however most have fallen by the wayside — a lot of them memorialized within the ‘NGS Necropolis’ (see go.nature.com/3xwvmkt). Catharine Aquino, who oversees short-read sequencing on the Purposeful Genomics Middle Zurich in Switzerland, attributes that success to technical experience. “It’s simply that the opposite corporations weren’t very dependable when it comes to library prep or sequencing,” she says.
Illumina’s portfolio consists of each compact benchtop methods for speedy evaluation of small numbers of samples, such because the iSeq Sequencing System that prices US$20,000, and the bigger high-end NovaSeq 6000, which prices practically $1 million however can churn out as much as 6 trillion bases (6 terabases) of sequence — roughly 2,000 instances the size of one of many human genome — each 2 days. Illumina’s new NovaSeq X household of production-scale sequencers use a redesigned circulate cell that may accommodate a a lot better density of sequencing reactions together with a retooled SBS chemistry and upgraded optics, in response to Illumina’s chief expertise officer, Alex Aravanis. The corporate studies that its new methods, which started delivery this yr, can generate as much as thrice as a lot information per run because the earlier era NovaSeq 6000, decreasing the price of sequencing to simply $200 per human genome.
An array of options
Past human genome meeting and mutational evaluation, new functions have fuelled demand for extra and higher short-read information at a decrease price. These embrace all the pieces from epigenetics to chromosomal conformation to proteomics. Aquino estimates that 60% of her facility’s work now includes single-cell RNA-seq, a sequencing-hungry approach that profiles the gene expression of hundreds or tens of millions of particular person cells. To fill that surge in demand, each start-ups and established corporations have entered the ring.
One established participant, MGI Tech, a Shenzhen-based spin-off firm from the Chinese language genomics titan BGI, provides distinctive twists on an Illumina-like SBS strategy. Each MGI and Illumina use a biochemical course of to generate a number of copies of each strand of template DNA on the flow-cell floor, thus boosting the fluorescent sign, however MGI’s DNBSEQ platforms use a lower-cost — albeit extra labour-intensive — methodology that converts templates into arrays of ‘DNA nanoballs’. “The info high quality is absolutely good, and it may be way more cost-effective than Illumina,” says Ioannis Ragoussis, head of genome sciences on the McGill Genome Centre in Montreal, Canada, who has used DNBSEQ devices in his personal facility.
Of the newcomers, the G4 benchtop system from Singular Genomics in San Diego might be most like Illumina’s. However the G4 additionally includes a flow-cell design that may make it simpler to run a number of sequencing experiments concurrently. “It’s actually focused in the direction of these smaller, extra versatile tasks,” says Stephanie Pond, vice-president of rising applied sciences on the Translational Genomics Analysis Institute in Phoenix, Arizona, which beta-tested the G4.
Ultima Genomics’ circulate cell is much more distinctive. Somewhat than utilizing a sealed cartridge containing advanced channels to coordinate the circulate of reagents, Ultima — based mostly in Newark, California — applies sequencing reagents to the uncovered floor of a spinning disc. The ensuing centrifugal drive distributes these supplies evenly throughout the disc’s floor, lowering each the complexity of the flow-cell design and the quantity of reagents required, and thus decreasing the price of every run. Ultima additionally cuts prices through the use of a combination of labelled and unlabelled nucleotides relatively than simply the more expensive labelled molecules1. In a single examine, early-access customers on the Broad Institute of MIT and Harvard in Cambridge, Massachusetts, documented usually comparable efficiency to Illumina in single-cell gene-expression experiments2.
Ultima Genomics’ devices run sequencing reactions on the floor of a spinning disc.Credit score: Carolyn Fong/New York Instances/Redux/eyevine
Lastly, there are the chemistries developed by Aspect Biosciences, based mostly in San Diego, and by Pacific Biosciences (PacBio)in Menlo Park, California, for brand spanking new short-read devices. Each depend on two-stage options to the usual SBS approaches, through which fluorescently labelled nucleotides aren’t completely integrated into the newly synthesized DNA however relatively bind transiently to the rising strand. As soon as they’re imaged, they’re then washed away and changed by unlabelled nucleotides.
This ends in a extra pure DNA synthesis course of whereas additionally permitting for cautious optimization of the labelling step, and each Aspect and PacBio — an organization already well-known for its refined long-read methods — spotlight the accuracy of their approaches. “We’ve been seeing extraordinarily high-quality information,” says genomics researcher Christopher Mason at Weill Cornell Drugs in New York Metropolis, who has used Aspect’s AVITI system to profile the consequences of area flight on human physiology.
Weighing up execs and cons
Sequencers fall broadly into two classes: production-scale devices together with Illumina’s NovaSeq, and smaller benchtop devices resembling Illumina’s NextSeq. For now, solely Illumina and MGI function throughout the complete spectrum; different short-read corporations goal particular ranges of throughput.
Manufacturing-scale devices are huge and costly, however such throughput is important for a lot of large-scale genomics or single-cell RNA-seq research, and such devices are inclined to kind the spine of core sequencing amenities. Stacey Gabriel, chief genomics officer on the Broad Institute says that just about all the sequencing achieved at her centre, one of many world’s main genomics amenities, makes use of such devices. “Now we have 32 NovaSeqs, and we run them very onerous,” she says, including that her crew might be augmenting this capability with new NovaSeq X devices.
Ultima additionally operates on this enviornment with its UG 100, however goals to counter the excessive price of its {hardware} with cheaper sequencing prices. The corporate claims that it has the potential to ship full human genome sequences for $100 — half the value of the NovaSeq X. The Broad Institute was one of many UG 100’s first customers, and Gabriel says that though the expertise remains to be maturing, she sees clear alternatives to include it into their workflow for whole-genome evaluation and high-throughput assays resembling single-cell transcriptomics.
“Now we have 32 NovaSeqs, and we run them very onerous,” says Stacey Gabriel, chief genomics officer on the Broad Institute of MIT and Harvard in Cambridge, Massachusetts.Credit score: Casey Atkins Pictures
With regards to buying selections, gear and reagents are solely a part of the calculation, and publicly introduced per-genome costs don’t account for labour, upkeep and different assist prices. Services can count on to pay 10% of an instrument’s base price yearly for service contracts, Ragoussis says, which may put even mid-range benchtop devices out of attain for a lot of labs. Most significantly, production-scale devices are solely less expensive relative to benchtop devices when they’re run at full capability. “There are quite a lot of tasks that simply aren’t large enough, or pilot-scale tasks the place it’s actually onerous to ‘feed the beast’,” says Pond. This will also be a problem for labs coping with a number of experiments that can’t be run concurrently in a single circulate cell.
Benchtop machines may be a greater match right here, and that is the realm through which PacBio, Singular and Aspect at the moment compete. Such devices usually price between $200,000 and $400,000, and there may be sturdy competitors to ship probably the most information on the lowest price-per-gigabase. “Price remains to be one of many greatest drivers, as a result of on the finish of the day individuals solely get a lot cash from grants,” says Mason. MGI has been utilizing this strain level to drive adoption of its merchandise, Mason provides, even by providing devices totally free to some labs which can be prepared to spend a set quantity on common orders of reagents.
Good software program untangles gene regulation in cells
High quality is one other essential consideration, and right here, too, Illumina has set a excessive bar. For many reads, Illumina’s methods will ‘name’ the right base 999 instances out of 1,000 — a regular of accuracy known as Q30 — and its newest-generation ‘XLEAP-SBS’ chemistry reportedly improves this accuracy by three-fold. PacBio claims that its new Onso instrument — which remains to be in beta testing — has error charges of 1 in 10,000 bases or decrease (Q40), and Mason says his check runs with validated genomic samples have borne this out. “In the beginning of the learn it’s even higher,” he says, reporting high quality practically an order of magnitude higher than Q40. Mason thinks additional optimization of the computational toolbox for analysing Onso-generated information may result in even higher efficiency.
A 2022 preprint3 from scientists at Aspect Biosciences additionally highlights the power to attain Q40 high quality for many bases in a human genome sequenced with the AVITI instrument, which began delivery in June final yr. The corporate additionally has a value edge over PacBio, matching Illumina’s $200 cost-per-human-genome and undercutting that of Onso by roughly sevenfold. In precept, greater high quality reads cut back the quantity of sequencing required for routine genomic research and will present a decisive benefit for functions such because the evaluation of circulating tumour-derived DNA in ‘liquid biopsy’ assays. “There’s comparatively few copies within the sea of regular DNA,” explains Gabriel, “so that you’ve obtained to sequence very deeply.”
One other consideration when selecting a sequencing platform is compatibility with current workflows. For instance, Aspect’s workflow is basically according to normal Illumina processes, whereas Ultima and MGI require further processing steps that may introduce pace bumps into current pipelines. “It’s not insurmountable — it simply provides extra time and labour,” says Mason. Additional automation may additionally be required to streamline the method.
Stability and reliability are additionally important, as a result of even temporary downtime can disrupt lab operations. Illumina usually has a wonderful status on this entrance, says Aquino. “Generally even earlier than we all know one thing is unsuitable, our engineer is there already,” she says. “It is going to take all these corporations a couple of extra years to construct up the assist system and this array of expertise.”
Going lengthy
Not each sequencing utility maps nicely onto short-read applied sciences. Subsequently, corporations resembling PacBio and Oxford Nanopore Applied sciences (ONT) in the UK have labored to evolve their long-read applied sciences as nicely.
Each corporations provide methods that immediately analyse particular person DNA molecules spanning tens and even lots of of hundreds of nucleotides. For PacBio, this entails feeding strands of template DNA into polymerase enzymes which can be tethered to a strong floor after which utilizing refined optics to detect the addition of particular person labelled nucleotides because the DNA synthesis proceeds. ONT methods decide nucleotide sequences on the premise of the distinctive adjustments in electrical present that happen as DNA strands transit by way of tiny protein pores. Collectively, these methods present insights that may be troublesome or inconceivable to acquire with short-read methods, together with massive structural variations in chromosomal DNA, mRNA transcript construction and full microbial genomes. Each methods also can immediately determine and map epigenetic modifications.
NatureTech hub
PacBio provides among the highest accuracy devices available on the market, because of a course of known as ‘HiFi’ through which the gadgets learn the identical phase of DNA time and again, ironing out random errors alongside the way in which. Nevertheless, they’ve traditionally been held again by excessive prices and low throughput. “100 samples in PacBio took a yr,whereas 100 samples in Illumina took perhaps two days,” says Aquino. However the firm’s new Revio instrument, which prices $779,000 and is scheduled to start delivery this yr, adjustments the equation. With the capability to attain 15-fold better throughput than current-generation methods, PacBio studies that the Revio can produce a high-quality human genome for simply $1,000.
ONT provides a uniquely versatile and moveable platform that may be simply as simply utilized to short-read functions as it may to ultra-long reads. Researchers usually use ONT methods within the area, and Mason has even despatched them to the Worldwide Area Station. “We are able to see functions in lots of distant areas,” he says. ONT additionally provides the lowest-cost sequencing {hardware} available on the market, together with the $1,000 MinION, which could be run off of a regular laptop computer or, in newer variations, a pill.
Against this, ONT’s high-performance PromethION can sequence as much as 14 terabases in 3 days, and makes use of an uncommon enterprise mannequin through which many of the upfront prices are related to the acquisition of consumables wanted to run sequencing experiments. “You get an instrument the place the value is related to what number of circulate cells you need to use with out you having to purchase it,” says Ragoussis, noting that this may be extra interesting than spending $300,000 or extra earlier than the lab even unpacks its first circulate cell. In October final yr, ONT launched its latest iteration of this platform, the moveable P2 Solo system, which may generate as much as two human genomes per flow-cell run and permits customers to get began for simply over $10,000.
In such a crowded market, the place change is a continuing, investing in new expertise requires a leap of religion. “It’s very troublesome to adapt each six months to a brand new expertise — it calls for loads from the neighborhood for benchmarking, for testing and in addition for the bioinformatics crew,” says Oliveira. For now, his crew is rigorously weighing up the professionals and cons of those rising platforms and the way they could complement or supplant his current {hardware}. However competitors, basically, is an effective factor, driving efficiency up and prices down. “We’re democratizing sequencing,” he says.