By Julia Karow
A year after Ion Torrent first presented its "post-light" semiconductor sequencing technology at the Advances in Genome Biology and Technology meeting in Marco Island, Fla., customers at this year's meeting talked about their first experiences with the Personal Genome Machine, and representatives from Ion Torrent and Life Technologies, the firm's new owner, provided a glimpse of where the platform is headed.
During a Life Tech conference workshop, Kevin McKernan, vice president of advanced research R&D, said that PGM customers can expect a 10-fold increase in output about every six months.
Since Life Tech launched the system in December, it has announced its first chip upgrade, from the Ion 314 Chip to the Ion 316 Chip. The upgrade promises to increase the output from 10 to 100 megabases per run and will be available to early-access customers this quarter, and more generally in the second quarter (IS 1/11/2011).
The best internal run today has yielded 300,000 reads 100 base pairs in length of quality Q17, according to Maneesh Jain, Ion Torrent's vice president of marketing and business development, who spoke during a separate Ion Torrent conference workshop. He said that the company plans to "address" RNA-seq as an additional application "later in the year."
Although the $500 price tag for the Ion 316 Chip is twice as much as that for the 314, reagent costs will stay at $250, so the cost per run will only increase from $500 to $750, and the cost per base will fall accordingly.
The system's read length is currently about 100 base pairs. According to information provided during the Life Tech and Ion Torrent workshops, read length is expected to increase to 200 base pairs in the fourth quarter and to 400 base pairs in 2012.
McKernan said that in a single run to sequence the E. coli genome, the system provided "uniform genome coverage regardless of GC content." Coverage of human genes has also been "very even" and has included areas that were missed by both SOLiD and Illumina sequencing, he said.
The PGM's per-base accuracy also continues to improve. According to McKernan, based on 50-base reads, it was about 98.7 percent at the end of 2010 and has since improved to 99.6 percent. In the second quarter of this year, it is projected to improve further, he added.
The company has also improved the accuracy for homopolymer regions, he said, largely based on better software. While the per base accuracy for a stretch of four identical bases was 94 percent at the end of last year, it has increased to 98 percent for five identical bases today, and is expected to go up to 99 percent during the second half of this year.
The sample prep workflow, which involves emulsion PCR for amplification, has been shortened as well. An experiment now takes about eight hours, McKernan said, including a six-hour sample prep process and a two-hour sequencing run. The sample prep can be done in batches of six, so up to six samples can be prepared and sequenced per eight-hour workday on two PGM machines, he said.
McKernan said that sample preparation previously took about 12 hours, which the company was able to reduce in part due to "new emulsion chemistries." Information provided during the workshops indicated that sample prep time is expected to shrink even further, to four hours in the third quarter and three hours by the end of the year.
Life Tech sells a benchtop instrument, the Ultra-Turrax tube drive, for use with the PGM to generate emulsion reactors of the right size, according to its website. Still, sample prep for the PGM appears to be largely a manual process at the moment. Life Tech is not selling its EZ Bead System, a three-instrument setup to automate emulsion PCR for the SOLiD system, for use with the PGM because the bead quantities needed for the PGM are much smaller than for the SOLiD, according to Mark Gardner, Life Tech's vice president and general manager of advanced genomic systems.
[ pagebreak ]
Asked whether Life Tech plans to develop an entirely new sample prep method that does not involve emulsion PCR, as suggested by some next-gen users, Gardner said that the Ion Torrent platform is "relatively open" and that "there are ways to get away from emulsion PCR."
"Whoever says that emulsion PCR is a barrier to entry for [using the PGM] will be proven wrong over time," he told In Sequence.
While Life Tech is promising improvements for the Ion Torrent platform, several customers reported their first impressions of the system during the conference.
At the Broad Institute, an early-access user of the technology since early last year, four PGM machines are currently moving "out of the implementation phase into the application phase," according to Chad Nusbaum, co-director of the Broad's genome sequencing and analysis program.
Data from the machines are "not a problem to use at all," he said.
The institute is currently getting approximately 20 megabases of data from each run, which includes 100-base reads of Q17 quality. It has just started to test the Ion 316 chips, which it expects to increase the yield five- to six-fold. "We typically run a couple of weeks behind” Ion Torrent, he said.
Nusbaum said he sees mutation validation, as well as quality-control of pooled libraries, for Illumina sequencing and of targeted capture samples as near-term applications for the system. As the platform improves, the institute is also considering it for applications like resequencing of microbes, exomes, and other targeted regions; viral sequencing; cDNA and transcriptome sequencing; and genome assembly.
Based on E. coli genome sequencing data, the platform seems to have a "very low GC bias," which he said is "surprising for a technology that requires an amplification step" and is "much better than what we have seen with other technologies."
There also was a "good representation" of homopolymers up to six bases in length, as well as "good performance" for insertions and substitutions. The number of falsely called deletions grew with longer homopolymers, but with new software, "more than half" of these went away, he said.
Stephan Schuster's lab at Penn State University received two Ion Torrent instruments at the end of October and has been "taking the PGM through its paces" since then, completing 52 runs since early November. Running the machines requires "very little hands-on time," he said, and his lab has so far used it to sequence and resequence the genomes of E. coli and Helicobacter pylori, which has a reputation for being difficult to sequence, as well as to analyze human mitochondria at high coverage for population genetic studies and to sequence short human amplicons.
He said that the majority of reads were around 100 bases long, and the data was "nicely distributed" across the bacterial genomes.
For H. pylori in particular, it is important to be able to distinguish between homopolymer runs of A that are five or six bases long, and he said the accuracy in those regions was better than 99.8 percent.
In human mitochondrial DNA, Schuster and his colleagues were able to call the majority of SNPs previously found by 454 sequencing, missing a few that were clustered in the hypervariable region of the mitochondrial genome.
They also sequenced a number of human PCR amplicons that were up to 20-fold mulitiplexed and pooled in a single run and were able to detect about three-quarters of previously discovered SNPs, a project that was completed in less than two days.
[ pagebreak ]
"The key application for this technology is that you can get results within a day," Schuster said. His lab is already performing four runs per day on its two instruments, and may be able to do six runs per day in the future.
Long Phi Le from the diagnostic molecular pathology laboratory at Massachusetts General Hospital said that his lab, which won an Ion Torrent PGM through the company's grant program last spring (IS 5/4/2010), received the instrument at the end of December and had it "up and running" within a week.
In 2009, the lab introduced a cancer genotyping assay based on SNaPshot technology, which assays more than 100 recurrent mutations in about a dozen cancer-related genes, but it is hoping to switch to next-generation sequencing in the future in order to capture additional mutations, structural variants, and gene expression levels.
They tested the PGM on six previously analyzed formalin-fixed paraffin-encoded cancer samples that were five to 10 years old, sequencing a number of PCR amplicons. Each run generated between 6 and more than 12 megabases, and between 85,000 and 102,000 reads. Of note, they were able to pick up all mutations except one that were detected by the SNaPshot assay.
The mean per-base error rate — using software that has since been updated by Life Tech — was about 1 percent, and the base substitution error was "very low," Le reported. "Yes, there is an issue with homopolymers," he said, adding that better base-calling algorithms will likely be able to improve that.
Right now, the Ion Torrent platform cannot yet fulfill Le's requirements for replacing the SNaPshot assay for analyzing FFPE samples: such an assay should provide 100 to 150 megabases of sequence data, or 200- to 500-fold coverage, to analyze between 100 and more than 1,000 exons, have a turnaround time of three to four weeks, and cost no more than $500 in consumables, he said. Ideally, it would be capable of sequencing an entire exome.
Joe Boland from the core genotyping facility of the National Cancer Institute said that his lab's PGM was installed in early January, and they have so far completed 24 runs.
He said the instrument is "easy to use" and it is "easily" possible to run it twice a day, though his lab is looking to run it up to four times per day in the future. The lab has generated between 17 and 28 megabases of data per run, and the read length is currently 100 bases, expected to grow to 200 bases by the end of the year, he said. The signal intensity has been "below spec", leading to higher-than-expected error rates, which his lab is currently investigating.
Boland and his colleagues want to use the PGM to sequence amplicons generated by Fluidigm's Access Array, which can perform up to 48 PCR reactions in 48 samples in less than a day. While it takes almost five days to complete an amplicon sequencing experiment with standard PCR and the 454 system, it takes less than two days with the Fluidigm and PGM technologies, he said.
To multiplex samples in a single sequencing run, the scientists have been testing a barcode design from the 454 system, with similar performance, he said.
Besides improving the signal intensity, he and his colleagues are working on standardizing the PCR reactions to obtain equal sequence yields; develop bi-directional assays; expand the barcoding scheme; optimize the amplicon length; and increase PCR multiplexing.
Regarding the emulsion PCR sample prep for the PGM, he said that the manual breaking of the emulsion "needs to be improved."
Baylor's Human Genome Sequencing Center received three Ion Torrent systems in early January and has completed and analyzed six runs to date, according to Donna Muzny, the center's director of operations.
The Baylor researchers have been able to obtain 21.4 megabases of data with Q20 quality per run, and Q20 reads with a mean read length of 80 bases. To evaluate the system's error rate and usefulness for genome assembly, they plan to test it on a variety of bacterial genomes with different GC content, as well as on mammalian BACs and amplicons.
Have topics you'd like to see covered in In Sequence? E-mail the editor at jkarow [at] genomeweb [.] com.