By Julia Karow
This article, originally published Feb. 17, has been updated with additional information from Oxford Nanopore and scientists.
MARCO ISLAND, Fla. - Oxford Nanopore Technologies said last week that it plans to commercialize two DNA strand sequencing instruments this year: a higher-throughput version that uses its GridIon platform and a disposable MinIon system that will be the size of a USB memory stick and cost less than $900.
The GridIon system, which takes disposable reagent cartridges that contain the nanopores, is intended to be scalable so that instruments can be combined in a similar fashion to computer clusters. Running 20 GridIon instruments, or “nodes,” in parallel will enable users to sequence a human genome at 15-fold coverage in 15 minutes for less than $10 per gigabase, the company said.
Clive Brown, the company’s chief technology officer, presented the technology at the Advances in Genome Biology and Technology meeting here.
At launch, both systems are expected to deliver read lengths of up to 100 kilobases with raw read error rates of about 1 percent. Company scientists have already sequenced DNA fragments up to about 50 kilobases in length and have achieved raw read errors as low as 4 percent.
In addition to the four DNA bases, the technology is able to identify methylated and hydroxymethylated cytosine and, in principle, it will be able to directly sequence RNA.
At launch, the GridIon system will use cartridges that contain an array chip with 2,000 nanopores and will produce tens of gigabases of data per day. The MinIon will contain about 500 nanopores and generate a couple of hundred megabases per hour.
Starting in early 2013, Oxford plans to increase the number of nanopores for the GridIon to more than 8,000, along with additional processing cores in each cartridge.
Both systems will produce data at a rate of between 20 and 400 bases per second and pore.
For either system, the run time will be user-defined - ranging from minutes to days - and will be limited by the lifetime of the electrodes, up to several days for the GridIon and about six hours for the MinIon.
Oxford Nanopore will have different pricing models for its platforms and consumables to accommodate different user budgets, offering discounts on either the instrument or the consumables. The firm said it plans to be competitive with other systems on the market on price per base.
For the GridIon system with 2,000 nanopores, for example, users can pay either full price on instrument and consumables, or get a discount on the instrument and sign a contract for consumables, or pay a higher price on the instrument and get a discount on cartridges. The cost per gigabase for that system is expected to range between $25 and $40.
For the GridIon system with 8,000 nanopores, available in 2013, the cost per cartridge will be higher, but the cost per gigabase will be on the order of $20 to $30.
The one-time-use MinIon, which plugs into an USB port, is expected to sell for between $500 and $900, depending on the number of instruments purchased. Software will cost a similar amount, but no additional IT equipment will be needed. The cost per gigabase on the MinIon will be on the order of $1,000.
The company said that it expects “further substantial pricing improvements” as the technology develops, in particular with increases in nanopore processing speed and denser electronic sensor chips.
Oxford’s strand sequencing technology uses an array of proprietary mutant protein nanopores that are embedded in a polymer membrane, which replaces the lipid bilayer it had been using in the past. The firm has designed a sensor array chip consisting of microwells that each have their own electrode. Application-specific integrated circuits, or ASICs, apply a potential across each nanopore and measure the ionic current flow. The sensor chip is contained in a cartridge that contains all the reagents needed for an experiment.
The company has not disclosed the nature of its nanopore, or whether it is derived from the alpha-hemolysin protein that it has been working with, but said that it screened 1,000 proteins and tested more than 300 mutants in detail.
The instruments take any type of double-stranded DNA with overhangs as input, which does not need to be highly purified because the polymer membrane is robust enough to handle “dirty” samples. For example, the system has been shown to be able to analyze DNA directly from blood or environmental water samples. Template DNA concentrations are typically in the picomolar range, according to the company.
Individual DNA strands are passed through each nanopore by a proprietary processive enzyme, which the company also does not disclose, though industry insiders speculate that it might be a helicase. The enzymatic DNA feeding technology was originally developed by Mark Akeson’s group at the University of California, Santa Cruz. He and his colleagues published an article last week on a system that uses phi29 DNA polymerase to feed DNA through a nanopore (see story, this issue), but the company said it is not using that enzyme. So far, the firm has not found a limit for the read length that is based on the processivity of the enzyme.
Oxford’s platform measures changes in the ion current through the nanopore that are caused by unique combinations of DNA bases, or "words," as they pass through a custom-engineered region inside the pore. Brown showed three-base combinations as an example but the firm is not disclosing the length of the "words" it reads, which would shed light on the pore it uses.
Base calling occurs on the instrument in real time. Each GridIon node contains computing hardware and control software for primary data analysis. Users can perform secondary analyses of the data while the experiment is ongoing, allowing them to continue the run until they have collected sufficient data to answer their experimental question, or to stop a run that looks like it is failing.
As a proof of concept, the company has, for example, sequenced the 48-kilobase genome of phage lambda as a single, contiguous read.
The sequencing accuracy does not deteriorate along the strand sequenced, and the majority of errors are deletion errors. Part of the error is systematic, meaning it can be fixed by further engineering the pore, Brown said.
Users will be able to stack several GridIon nodes, and Oxford plans to sell them as bundles ranging from a single node to up to 20 nodes.
A 20-node installation, using 8,000-nanopore cartridges, is expected to deliver a complete human genome at 50-fold coverage in 15 minutes, according to the company, or 3 terabases of data per day, based on a sequencing speed of 300 bases per second. For that setup, the cost per gigabase is expected to be under $10.
Oxford Nanopore plans to announce about a dozen or so early-access customers for both the GridIon and the MinIon in the near future, and expects that these customers will talk about their experience with the systems in the fall.
The company is not yet taking orders for the platforms but plans to launch the systems commercially during the second half of this year.
Reactions to Brown's presentation at the conference were generally characterized by excitement — especially regarding the tiny MinIon platform — and cautious optimism, mixed with occasional disbelief that the company will be able to deliver on its claims. Most scientists said they need to see more data, including details on the raw signal the company collects, to be able to judge the technology.
"It exceeds my expectation — at least their claims exceed my expectation," said Chad Nusbaum, co-director of the genome sequencing and analysis program at the Broad Institute. "And if it's true, it certainly will create some excitement in the field, and the other vendors are going to have to react." Nusbaum added that he would like to test one of the instruments "as soon as I can."
"I know from talking to multiple people here that they are very excited about the possibilities of that technology," said Elaine Mardis, co-director and director of technology development at the Genome Institute at Washington University, noting that Brown's presentation "caused a lot of stir" both at the conference and in the electronic media.
According to Harold Swerdlow, head of sequencing technology at the Wellcome Trust Sanger Institute, it is "a little too early" to comment on the technology in detail. The fact that the platform requires no sample preparation and could potentially sequence DNA directly from blood is "incredibly exciting," he said, for example for malaria projects conducted by Sanger researchers in Africa.
Have topics you'd like to see covered in In Sequence? Contact the editor at jkarow [at] genomeweb [.] com.