By Julia Karow
Ion Torrent has released initial performance specs for its Personal Genome Machine and an early-access user from the Broad Institute has provided his first feedback on the platform.
According to specifications released at a Life Technologies-organized meeting last week, the PGM will initially generate at least 100,000 single reads per run with a read length of 100 to 200 base pairs and an error rate of about 1 percent, using the so-called 314 chip, which has about 1.4 million sensors and costs $250. The company plans to start shipping the instrument worldwide later this month.
As previously reported, sample preparation takes about a day — at least 10 hours — and the instrument will cost $49,500, not including a $16,500 Ion Torrent-branded Dell server to analyze the data. An upgraded chip, the 316, which will have about 6 million sensors and cost about twice as much, is planned for next year.
Internally, the company has been getting runs with close to 200,000 reads 100 to 200 bases in length, and 400,000 reads 50 bases in length, Jonathan Rothberg, the company's founder, CEO, and chairman, told In Sequence this week. Company researchers have obtained more than 40 megabases of high-quality data in a single run, and have already achieved reads longer than 300 bases.
"I have never worked on a technology that moved as fast," said Rothberg, who founded Curagen, 454 Life Sciences, and RainDance Technologies prior to Ion Torrent.
Life Tech has been taking orders for the PGM for several weeks now and, in preparation for shipping the instrument worldwide later this month, has stocked up its distribution warehouses, Rothberg said, so instruments can be delivered before the end of the year.
Because the instruments and chips are mass-produced in electronics equipment factories, the company does not expect to be constrained by manufacturing to fill its orders. "We can ship chips in unlimited quantity from day zero, and machines in unlimited quantity from day zero," Rothberg said.
Tech support is also in place — one of the reasons Ion Torrent teamed up with Life Technologies, Rothberg said. "As long as we operate at the limit of modern electronics, nobody can beat us from a technical point of view. Then it becomes all about support," he said, adding that Life Tech has field service engineers and application specialists worldwide "that we are leveraging from the start." In addition, Life Tech can provide reagent kits for a variety of applications.
While there will be upgrades to the chips and instrument software starting next year, no hardware changes will be necessary to increase the instrument's performance. "When the chip is the machine, it's very easy for us to upgrade," Rothberg said.
About 80 percent of potential customers Ion Torrent has talked to are interested in targeted or amplicon sequencing of gene panels, and given the platform's current throughput, "we're a perfect match for that," he said. Potential applications, he added, are in genetics of inherited disorders, cancer research, and viral research.
As the throughput of the Ion Torrent sequencing chips increases over time, the platform will become suitable for exome sequencing. "And of course, in the next three or five years, everyone wants to do whole genome, and we'll be there with the chip," he predicted. Rothberg did not mention whether or when the company plans to introduce paired-end sequencing.
At the moment, Ion Torrent is not betting on any particular technology for enrichment of targeted sequences. There are at least five methods available now, he said, including multiplexed PCR, RainDance, Fluidigm, NimbleGen, and Agilent, and he sees "not one winning it all yet."
[ pagebreak ]
A User Weighs In
As of now, Ion Torrent machines are in the hands of at least six early-access customers across the US, most of them undisclosed. Several are expected to present data at the Advances in Genome Biology and Technology meeting in early February, Rothberg said.
One early-access user is the Broad Institute, which has been collaborating with Ion Torrent since earlier this year on optimizing laboratory processes and on developing applications for the platform. The institute currently has four beta instruments on site.
Broad researchers are getting "similar performance" to Ion Torrent's launch specs, Chad Nusbaum, co-director of the institute's genome sequencing and analysis program, told In Sequence this week, adding that "we're running a little bit behind them."
According to Nusbaum, who also spoke at the Life Tech meeting last week, the platform is "sort of a tactical machine" that is "easy to roll out and easy to change what you're doing with it." Initial advantages lie in its speed, the low cost of the instrument, and the ease of handling consumables, he said, noting that the platform cannot yet compete with currently available high-throughput sequencing platforms in terms of cost per base.
Overall, he said, the machine is "relatively easy to operate and easy to implement" and the "computational overhead isn't too bad." In addition, because the reagents and enzymes are generic and commercially available, and the Ion Torrent chips are "imperishable," quality control and storage of consumables are straightforward.
One area where speed is a big plus is in research and development, Nusbaum explained. "If it takes you a week and a half to do your sequencing, then your development cycle — if you use sequencing as your readout — is stuck at one and a half weeks. But if you can have your sequencing readout in an afternoon, you can do many more cycles of tech dev."
Since the system's output of 5 to 10 megabases per run is not yet enough for human whole-genome sequencing, the Broad is testing it for targeted sequencing applications. For clinical or medical sequencing in particular, "it's really nice to be able to go fast," he said.
The institute is also exploring the platform for bacterial genome sequencing, though speed might not be of the essence for that application. "It's better for us to go cheap," Nusbaum said.
Another appealing aspect of the platform is its low price, he said, because it frees users from having to run their instrument all the time to make it worth the purchase, as with an expensive instrument.
"I think there is a large number of groups who could get their hands on the $50,000 required to buy this machine with not too much difficulty and then would be able to run it as required," he said. "They are not going to use it every day, but having it would allow them to have control over their sequencing. And that's a big goal of a lot of people."
Have topics you'd like to see covered by In Sequence? Email the editor at jkarow [at] genomeweb [.] com.