NEW YORK (GenomeWeb) – Oxford Nanopore Technologies last week started broadly offering a flow cell adapter for its MinIon and GridIon sequencing platforms that uses small, disposable flow cells and allows users to perform small-scale experiments at a low cost.
The adapter, called Flongle ("flow cell dongle"), which has been in the works for several years and has been tested by early-access customers over the last few months, promises to facilitate nanopore methods development, which could result in new applications, including clinical tests that may run on the platform. Combined with planned sample preparation devices and the integration of computing, it could also broaden the use of nanopore technology outside the laboratory.
At the moment, customers can only purchase Flongle as a $5,250 package that includes the adapter and 48 Flongle flow cells and requires users to have a MinIon or a GridIon. However, the company plans to start selling $90 flow cells soon, in packs of two, and is considering other packages that would combine Flongle with the MinIon, GridIon, or the forthcoming MinIon Mk1C device.
Flongle flow cells, which are one-time-use only, currently deliver up to 1.8 Gb of sequence data and might generate more than 3 Gb in the future, according to the firm, though most early-access users have produced less data in their experiments.
According to Oxford Nanopore, interest in the adapter has been very strong prior to the release, and the firm is currently scaling up production of Flongle flow cells. It did not say, though, what fraction of existing MinIon and GridIon customers it expects to purchase the adapter.
The company also cautioned on its website that it anticipates "very high demand on this product" and will "endeavor to meet your shipping schedule," implying that it may not be able to fill Flongle orders immediately.
Early-access users said the data from Flongle flow cells is very similar to that of other flow cells, but the yield is lower, since Flongle flow cells only have 126 channels compared to a MinIon flow cell's 512 channels.
They also pointed out that Flongle flow cells are more user friendly than other nanopore flow cells because they don't require priming — several pipetting steps that can potentially introduce errors. Oxford Nanopore has said it plans to make other flow cells priming-free in the future but has not provided a timeline yet.
Several early-access customers plan to use Flongle flow cells to develop and test new nanopore assays and methods, since the cost is much lower than that of a MinIon flow cell and there is no need for multiplexing samples to save money.
William Jeck, a fellow in gastrointestinal pathology at Brigham and Women's Hospital, for example, has been developing a gene fusion detection assay for sarcomas and hematologic malignancies on the MinIon that he hopes to deploy clinically in the future. He has tried the assay on four Flongle flow cells so far and believes they could potentially be used for clinical applications. "The major benefit is that it's incredibly fast to perform a run, and the per-sample cost is really quite low," he said. In addition, samples no longer need to be batched, so they can be processed immediately.
Previously, he had to multiplex four samples on a MinIon flow cell to keep costs down, but now he can run a single sample on a Flongle flow cell at half the price while getting at least the same amount of data. "So, for us, it's a no brainer," he said.
His assay, which relies on amplicon sequencing, does not require a lot of data, which is why he has not cared much about the yield per Flongle flow cell. Rather, he welcomes the opportunity to develop the test faster. "We haven’t been able to do it until now because sequencing costs $1,000 per [MinIon] flow cell, so, iterating on that machine was expensive. Iterating on a $100 flow cell is suddenly much more feasible. Our ability to test assay preparation techniques is dramatically increased, and we're actively in the process of doing that right now," he said.
Other early-access users also plan to employ Flongle for methods development. Matt Loose, an associate professor in the School of Life Science at the University of Nottingham, for example, has tested about 20 Flongle flow cells in total so far and has used some to sequence samples prepared with new long-read DNA extraction protocols, which his lab has been developing as part of a group called Long Read Club. "We don't need to sequence tens of gigabases in order to see how our method is going," he said.
In addition, he has used Flongle to test a CRISPR-Cas9 target capture protocol for sequencing a region of the human genome, generating 100 Mb of data, or 50x coverage. "We don't actually need to run a full MinIon flow cell for that type of project," said Loose, who runs a sequencing facility called Deep Seq that has Oxford Nanopore's PromethIon, GridIon, and MinIon installed.
Going forward, he plans to use the Flongle for targeted human sequencing and will look into "how we can deploy it to non-traditional sequencing environments," he said.
In runs that didn't focus on a specific target or obtaining ultra-long reads, he said, his lab has had yields between 600 Mb and 1 Gb per Flongle flow cell. "In the early-access phase of any of these technologies, sometimes the chips don't have as many [active] pores as you'd like," he said. "But even on those, we've been able to do things like sequence long-range PCR products, where just getting a couple of thousand reads is sufficient to answer the question. So, for something where you're not expecting huge amounts of data, they're really good."
He and other early-access users have also found Flongle to be useful for quality control of sequencing libraries before they go onto expensive MinIon, GridIon, or PromethIon runs. As a sequencing service center for PromethIon, Loose's lab often obtains samples of unknown quality, and checking the library on a Flongle flow cell "gives us an early look if there's a problem with the DNA," he said. His facility plans to build this into the PromethIon service offering, he added.
Loose said he also found that yields from a Flongle run tend to predict the yield one can get from bigger flow cells. "If you take the same sample and run it on Flongle, if you hit 1 Gb, you'd expect 10 to 15 Gb on a MinIon and probably more than 100 Gb on a PromethIon," he said, "whereas if you only got 100 Mb on your Flongle, that sample is going to give you proportionately less on MinIon, GridIon, and PromethIon. It scales quite well in our admittedly fairly limited experience so far."
Mark Akeson, a professor of biomolecular engineering at the University California Santa Cruz, agreed that checking nanopore sequencing libraries is a good use case for the Flongle. "There was no unambiguous way prior to the Flongle to test the quality of the library before you put it on the flow cell," he said.
Akeson, who has been consulting for Oxford Nanopore for more than 10 years, said his lab routinely obtains 500 to 700 Mb of data from a Flongle flow cell with genomic DNA, and the number of active channels has ranged from 50 to 75.
He also thinks that Flongle will be ideal as a teaching tool, for example, to train students on nanopore sequencing technology before they go on to use expensive MinIon or PromethIon flow cells. "A student can make an attempt on Flongle half a dozen times" at a cost lower than that of a single MinIon flow cell, he said. "You can get more people in the lab working on the device, because they can learn."
"We could use all the Flongles that Oxford Nanopore sends us," he said. "The demand is really high, so it's been hard to get them."
John Tyson, a senior research associate at the Michael Smith Laboratories at the University of British Columbia, also believes Flongle will come in handy for teaching and training, both of researchers and non-scientists. His department has a teaching laboratory, he said, that has already trained scientists on nanopore sequencing and sometimes brings in high school students for molecular biology classes. "Flongle takes it to a price point where people are happy to spend this sort of money for high school groups," he said. "I think we will see a lot more of that appeal to the education market."
Tyson's lab, which has four MinIon sequencers, has tested four Flongle flow cells so far, using them for verifying clone sequences and for amplicon sequencing to look for splice variants. Yield has varied from 600 Mb to a record 1.8 Gb so far. That best run was an amplicon sequencing experiment and used a sample that previously performed well on a MinIon flow cell, he said. As with other nanopore flow cells, yield will depend on sample quality and the specific application, he added.
Like others, he plans to use the Flongle for methods development going forward. "Because it's so cheap compared to, say, a MinIon flow cell, you're willing to take more of a risk with it," he said.
In addition, he could imagine Flongle becoming a standard molecular biology tool — like a benchtop centrifuge or a vortex machine — that researchers can use to verify sequences of clones quickly, rather than sending them out to a sequencing facility, like his lab currently does. "I can do a rapid prep and have sequencing come out in probably 5 to 10 minutes. And as long as I don't need huge amounts of data, I’m just confirming a clone, it's going to be perfect for that," he said.
He and others also look forward to the price of Flongle adapters and flow cells coming down further. "At the moment, it's probably still a bit too expensive," said Loose. "I'd like to see it cheaper. It's the sort of thing that could be deployed in the field very easily."