This story has been updated to include additional information from Oxford Nanopore.
NEW YORK – Oxford Nanopore Technologies said on Wednesday that it plans to release a new chemistry as well as several new pieces of hardware over the next year.
The firm also recently signed a $68 million contract to expand a sequencing project in the United Arab Emirates and, as a result, has raised its financial guidance for 2022 and 2023.
The announcements are the first product and business updates since the company went public on the London Stock Exchange in September.
In its prospectus, Oxford Nanopore had said it expected 2023 revenues in the range of £165 million ($219.3 million) to £175 million. But following the award of the three-year $68 million contract with G42 Laboratory last month, the firm raised its 2023 revenue guidance to between £170 million and £190 million. It also introduced revenue guidance for 2022 of between £135 million and £145 million. The firm is maintaining its 2021 revenue guidance in the range of £105 million to £111, as it does not expect to generate much revenue from the deal this year.
In a technology-focused talk at the firm's user meeting, held virtually this week, Clive Brown, Oxford Nanopore's chief technology officer, and other company officials highlighted ongoing R&D efforts.
The PromethIon 2 (P2) platform, first disclosed in May, will come in two flavors. The standard P2, with instruments set to ship by Q3 next year for $59,955 per device, offers onboard computing with a graphics processing unit, while the P2 Solo, a smaller instrument that will cost $10,455, will need to be connected to a computer or a device with onboard computing, such as a GridIon. The P2 Solo will be available in Q2 2022. Consumables pricing should enable human whole-genome sequencing for under $1,000 a sample, Brown said. The firm is also working on an update to the MinIon Mk1C portable sequencing platform that will be based around the Apple iPad Pro.
"The P2 Solo really enables users to bridge the gap between the MinIon and the PromethIon 24/P48," Rasmus Kirkegaard, a researcher at the University of Vienna and a customer of Oxford Nanopore, said in an email. Researchers that occasionally need to generate large amounts of sequence data per sample, but don't run enough samples to justify the larger instrument, may find it the most useful, he suggested.
With sequencing output of up to about 600 Gb, the P2 Solo "is quite competitive with other setups," Chris Mason, a sequencing expert at Weill-Cornell Medicine, said in an email. "We will buy one and check it out."
Oxford Nanopore also touted its work on improving accuracy, through its Q20+ kits, now called Kit 12, which feature new pore and motor proteins, as well as a longer-term project to enable its devices to re-sequence single molecules to reach whatever consensus accuracy a user desires. This ability could enable liquid biopsy applications, for which the firm is developing a new flow cell that can accept raw samples.
Though the Kit 12 and P2 instrument had been announced prior to the IPO, Brown provided more details on the releases.
The improvement to raw read accuracy is enabled by the new R10.4 flow cell and e8.1 motor protein, as well as software improvements. The kits provide better accuracy but are "a tad slower" than the current ones, Brown said, noting that the speed through the pore has been dialed down to about 250 bases per second. However, sequencing speed depends on the particular basecalling algorithm used.
The kits can make use of "duplex" reads, where complementary strands of DNA are sequenced one right after the other. The modal accuracy of duplex reads sits around Q30, Brown said.
Customers have seen similar results. "For maize, we generated duplex reads up to ~100 kb in length that showed a percentage alignment identity of 99.94 percent when aligned to the golden standard reference genome," said Alexander Wittenberg, a scientist at Dutch ag-bio company KeyGene, in an email. He noted that even using single reads, the new chemistry allows "spectacular, near telomere-to-telomere assemblies," using a plant-specific basecaller. "We obtained assemblies that were on par or even surpassing the [Pacific Biosciences] HiFi-based assemblies in terms of contiguity and consensus accuracy," he said.
Longer term, Oxford Nanopore is also seeking to improve accuracy by providing more than two reads per molecule. The company has circled back to a nanopore and motor protein configuration that ratchets molecules "out" of the protein, against the direction of the electric field. This so-called "outy" chemistry allows the sequencing device to bring the nucleic acid back through the pore to be sequenced again. "It turns out we can reread a single molecule again and again and again," Brown said, essentially as many times as needed. The strategy would also allow the device to size the molecule before sequencing, and it could even be told to sequence only molecules of a certain size.
Already, using an unoptimized chemistry, three reads of the same molecule can generate a median accuracy of 99.5 percent. "You can be basecalling and assessing until you've hit the accuracy you want," Brown said. Combined with adaptive sampling, which can eject molecules from the nanopore that don't match a predefined sequence, a researcher could scan for very rare sequences, like a cancer driver mutation or the genome of a rare animal in an ecological context.
This approach could be used in liquid biopsy, an application that Oxford Nanopore is also developing new hardware for. Brown said he hopes to have a prototype flow cell for this purpose by the "middle of next year." The flow cell would accept a raw sample as input, and DNA would be moved to the detection chamber by electrophoresis. Internally, the firm is using a modified MinIon flow cell.
Brown also said the firm is developing a basecaller that runs in a web browser such as Google's Chrome. The company is also working on the MinIon Mk1D portable sequencing platform. Like the Mk1C, it would consist of a case that integrates a MinIon flow cell. But using the web-based basecaller and 5G connectivity, it could be run just about anywhere. Using the iPad Pro, which runs on Apple's new proprietary microprocessors, "we're getting 100 kb per second," Brown said, which is "close to P2-level performance." The Mk1D could be available before the end of 2022, he added.
According to Wittenberg, the Mk1D could make sequencing even more ubiquitous. "Like the Mk1C, it will allow sequencing and analysis outside a laboratory environment (e.g., in the greenhouse or in the field)," he said.
Elsewhere, the firm is continuing to add to the ability to call methylated bases. Already able to call 5mC and 5hmC modifications, Oxford Nanopore is in the process of adding 6mA followed by 5caC, 5fC, and 4mC. It is also adding the capability to sequence short fragments for use in counting applications to its MinKnow operating software.