Oxford Nanopore Announces New Pores, Kits and Updates on Projects
By Allison Proffitt
September 29, 2016 | Clive Brown, Chief Technology Officer at Oxford Nanopore (ONT), gave a technical Oxford Nanopore update this morning, outlining updates on MinION, PromethION, SmidgeION, hardware and chemistries.
Billed as a “wafer-thin update”—Brown is a documented Monty Python fan—the webcast was instead chock full of updates.
Brown began the webcast by expressing his excitement over NASA’s recent use of the MinION in their breakthrough genetic sequencing in space.
Brown continued by announcing an update to the protein pore—version R9.4. The updated pore will offer faster, more accurate sequencing. In the future, Brown stressed, all upgrades will be backwardly comparability with the R9 pore. The enzyme, or motor, of the nanopore sequencing setup will also start shipping updates in October: E8 will be able to improve speed from 250 bases/second to 450 bases/second without a drop in accuracy, Brown said.
Along with some adapter updates, the new pore and motor in MinION Flow Cells now consistently yield >5Gb in the first 24 hours.
There have also been updates to the flow cell itself. The plastic was recently updated, Brown, said, though “nobody noticed,” and an input pore was added above the nanopore array. From the beginning, Brown said, the goal was to create a flow cell that could be shipped and used at ambient temperature. Thus far, Oxford Nanopore has shipped via a cold chain, but the new SpotOn Flow Cells will be easier to use with a longer shelf life.
Getting rid of the cold chain was a recurring theme. Brown announced a rapid click chemistry kit needing no external reagents. “There should be field-deployable, non-cold chain versions of the sample prep as well, coming on quite shortly.”
Brown announced more MinION kit changes as well. Though ONT will continue producing both 1D and 2D kits, he said he hopes to push users toward 1D kits, making it “the defacto standard as it gets better and better.”
The new rapid 1D kit will reduce library prep to 5 minutes (from 10) and require no additional reagents. With the updates, users should be able to get a 1D mean accuracy above 90%, Brown said, and he expects to be able to improve that. (There won’t a new rapid 2D kit.)
Brown announced (and editorialized on) several updated kits. There will also be a new wash kit for the MinION to flush old samples out of the flow cells; a new short fragment kit (“Why anybody would want short DNA? I don’t know.”); a new rapid PCR barcode kit, a new rapid PCR kit, and a new rapid amplicon kit. The company is planning to make a direct RNA kit available to the developer group. Every kit also comes with multiplex versions.
On the data analysis side, MinKNOW sees continuous improvements. There are now 3 ways to basecall in real time: live & local, local & offline, and in the cloud, though Brown hopes to “phase out” base calling in the cloud; aiming for a November 5 date.
He said the company hopes to move base calling to FPGAs (field programmable gate arrays). FPGAs are more power efficient than GPUs or CPUs; researchers have been able to get about 80% utilization of the FPGA. It takes only 3-5 “state of the art” FPGAs to keep up with a PromethION, Brown said, and a single FPGA can keep up with the MinION. Brown said the company plans to develop an “FPGA on a stick” that will do “very fast real time base calling to keep up with the increased speeds.” Initially, he expects an FPGA on a separate memory stick for the MinION, though he hopes to integrate soon. The PromethION FPGAs are already integrated.
Brown also announced Albacore, which will be available to customers and developers on October 16. “Any future major developments and innovations in base calling will be available through this Albacore pipeline,” Brown said.
All of these MinION updates—to the flowcells, kits, new scripts and base caller upgrades—will be available for order on October 10, shipping October 17.
Brown highlighted VolTRAX, a device announced at ONT’s event in May. VolTRAX is intended for clinics, doctor’s offices, and field hospitals, Brown said, “the kind of in between places where there is some prospect of somebody wielding a pipette”.
Brown showed the VolTRAX, a programmable “base station” with integrated heating and cooling connected to a lap top onto which a single use flow cells with multiple sample and reagent ports clicks. “The VolTRAX can do the standard preps, of course,” Brown said, “but we see it doing much more complicated preps in the future.”
“We are in the process of adapting and optimizing all of our current library preparation protocols for use on VolTRAX, and currently have ligation-based 1D and 2D library preparation protocols, as well as our rapid 1D protocol, working well on the device,” the company says on the VolTRAX data page.
The access program for the VolTRAX opens on October 11, Brown said. First devices will be shipped around the time of the Nanopore Community Meeting in early December.
The PromethION access program is currently open. Brown declined to tell how many individuals are on the waiting list for access, but said there were many. The PromethION Access Program will likely close this year, he said.
He shared the first actual photo of a SmidgION, the mobile-phone compatible DNA sequencer. (In May, mockups of the SmidgION showed a device that attached to an iPhone with both the lightning and headphone ports. The mockups Brown showed today only had a lightning port. The actual device was only shown connected to a phone.)
“Almost all of the electronics are in this box, in the sensor box,” Brown said. SmidgION will support R9 chemistry and will be capable of 450 bases/second.
“It’s really targeted at mobile testing,” Brown said. “If you imagine a SmidgION with an embedded Zumbador-type device, and embedded targeted chemistry, again it becomes a amenable package for distributed testing… With that speed you can do a lot of rapid testing.”
Brown also gave insight into a few of the projects on which ONT is currently working.
We’re seeing more of a demand to focus on a subset of input materials, he said. With portable devices, you can focus more and more on low-cost, high volume testing. Brown described nanopore sequencing as a platform on which to expand discovery, translation, and deployment of tests, both clinical and nonclinical. “Testing food or the environment, it doesn’t matter. It’s all the same to us,” he said.
His first example has already been published by MinION users. Matthew Loose and colleagues published in Nature Methods in July (doi:10.1038/nmeth.3930) detailing real time selective sequencing. “In real time you can enrich your fragments of interest and just sequence those,” Brown said. “You can pull out a bit of the human genome, you can do barcode balancing—that’s quite a cool one, I think, for amplicons.”
Brown also summarized Oxford Nanopore’s own experiments with Cas9, using inactive protein to “call-down” select regions of DNA as part of sample prep, hinting that the company would soon offer standard test kits for sale.
He showed an example of Project Zumbador, the low cost, reproducible sample and library prep device he introduced in May that ONT intends for lab-free work in the field. Sample input can be, “anything gloopy”, Brown said—blood, saliva, wound pus, or food—and “within some number of minutes you’ll be able to sequence what’s in there.”
Zumbador is getting simpler and simpler all the time, Brown said. ONT has begun using a bead capture and delivery system. Dehydrated beads sit at the bottom of the Zumbador device. “Once the beads become hydrated [with sample], they drop directly onto the flow cell,” Brown said. “It’ll probably have to have a shippable liquid phase… that contains a cell lysis soup and then some kind of push activation that allows that to wick into the dry layer, the sample prep to occur, and then after a few minutes, the beads drop onto the flow cell.”
The bit closest to the flow cell is the most mature, Brown said, while sample input is still being developed.
Farther out, he announced a Field Effect Transistor-based million channel sensor that he says the company has been working on for “quite a while.” The FET array will still be a small-chip—the same form factor as the MinION or PromethION—Brown said, but the sensors will be, “packed in at very high density.” So far, the FET arrays have shown very good signal to noise ratios. In principle, Brown theorized, you could have a million channels running 1,000 bases/second.
He called the FET arrays “next, next big technology shift,” stressing that the while the proof of concept has been completed, there’s much research left to do and a product is still at least two years away. “But it is basically working!”
He teased a solid state pore, saying the company can now repeatedly measure DNA passing through solid state holes, and generate “the same shape signal for the same sequence of DNA.” Functioning solid state sensors would be put on the FET array, he said. “It’s not clear whether solid state holes will be superior to proteins—I’m not sure they will be in terms of the quality of the signal—but there may be other practical reasons to do solid state holes.”
Also in the pipeline, Brown mentioned a next generation ASIC, and a “Flongle” or flow cell dongle that was initially an internal test device that ONT has decided to make available.