By Julia Karow
This article, originally published April 28, has been updated with additional information from the two winning teams.
Research teams at Massachusetts General Hospital and the Marine Biological Laboratory at Woods Hole who each won an Ion Torrent Systems sequencer through the company' grant program last week plan to use their respective instruments in cancer molecular diagnostics and water quality monitoring projects.
Both teams said they are intrigued by the low cost of sequencing and fast turnaround time that the new platform will offer, though the jury on data quality is still out.
Sometime later this year, when the platform is ready for commercial release, each team will receive an Ion Personal Genome Machine sequencer, along with subscriptions to DNAStar's SeqMan NGen assembly software and CLC Bio's Genomics Workbench bioinformatics software.
According to Ion Torrent CEO Jonathan Rothberg, the company received "hundreds of grant applications" and chose the winners because they offered "innovative solutions" for human healthcare and the environment.
Cancer Genotyping at MGH
One of the awards went to John Iafrate, director of MGH's Diagnostic Molecular Pathology Laboratory, and Long Le, a clinical pathology resident and clinical fellow in Iafrate's lab. They are planning to use the sequencer eventually in routine clinical care, for targeted genotyping of tumors from a variety of cancer types.
Last year, their lab introduced a multiplex single-base extension cancer genotyping assay, called SNaPshot, which analyzes more than 100 recurrent mutations in 19 cancer-related genes using multiplex SNP detection technology from Life Technologies. So far, the team has analyzed more than 1,000 tumor samples using this assay — mostly lung and colon cancers — but a sequencing-based assay would offer some advantages.
"We are getting a fair amount of interesting information with this clinical test, but the limitations are quite significant," Iafrate told In Sequence last week. "In large genes, especially tumor suppressor genes, we really don't have the capability of uncovering possibly clinically very important mutations."
His group's aim is to develop a sequencing-based assay that would initially cover between 500 and 1,000 exons. Within six months or so, they are planning to narrow down the list of targets and explore enrichment strategies, both those based on multiplex PCR and on DNA capture. Within a year or so, they want to validate the assay and sequencing system, so a year and a half to two years from now, there will be "a complete assay, ready to launch," Le said.
He and his colleagues have already started pilot projects to test several sequencing platforms — including Illumina, SOLiD, and Helicos — for their ability to analyze poorly preserved DNA from formalin-fixed paraffin-embedded samples, and are comparing results to those from the SNaPshot assay.
Sometimes only 10 percent of a biopsy contains tumor cells, so the assay needs to be very sensitive, Le explained. Next-gen sequencing platforms can achieve that through higher coverage, he said. According to their calculations, at least a 200- to 500-fold coverage is required to reach their target sensitivity of 5 percent.
While Ion Torrent's sequencer is just one of several platforms they plan to test, they are already intrigued by several of its attributes. "If it works comparably to the other platforms in terms of error rate, I think there is no doubt that we would put significant resources on Ion Torrent," Iafrate said. "All the other advantages are so significant that it makes it quite exciting."
[ pagebreak ]
About six months after receiving the instrument, he said, "we will have a very good idea of how it compares to other platforms."
One big advantage is the fast run time of an hour or two, "which is ideal for clinical applications," Le said, because in the case of failure, a run could be repeated the same day. Runs might even be faster than in conventional Sanger sequencing, he added.
Also, the $50,000 instrument price — about an order of magnitude lower than that of other sequencing systems — "is something we can afford, as a clinical lab," he said, and it would not be difficult to scale up operations if needed.
The instrument's small footprint — it fits on a laboratory bench — is also "very attractive" compared to other next-gen sequencing instruments, he said. "To be honest, right now, if we were to get a SOLiD system, I don't even know where we would put it, because we don't have that space," he said.
And because the Ion Torrent platform does not generate image data, "it bypasses the need to have large amounts of server space and data storage space," according to Le. "This was very attractive to us because we are not a huge lab; we're not a genome center … so we really need a simple solution in terms of the data." Having a license to CLC Bio's software will also help, he added.
Initially, Ion Torrent said, the instrument will generate read lengths of about 100 bases and 100 megabases of data per run at a run cost of about $500.
While that might not suit every need — for example, human whole-genome sequencing — "for our application, it really is perfect," Iafrate said.
And though additional molecular information about a cancer might be useful — such as microRNA expression, or epigenetics — a targeted cancer genome sequencing assay might last for a long time. "We really think we can provide almost all of the clinically relevant information with a single assay — copy number, translocations, mutations, insertions, deletions. We're really excited that this may be the test of choice for the foreseeable future."
Testing the Waters at MBL
Ion Torrent's second grant went to Mitchell Sogin, director of the Josephine Bay Paul Center for Comparative Evolution at the MBL in Woods Hole, Mass. He and his colleagues plan to use the instrument to monitor water samples in order to identify types and sources of fecal contamination.
Using their in-house 454 sequencing system, they have already developed a method for profiling complex microbial communities, which they want to apply to monitoring environmental waters for pollution with human sewage.
Traditionally, he said, scientists have used indicator organisms, such as E. coli, to determine whether sewage has been released into the water, but that can be misleading, because these bacteria can also come from other sources, such as vertebrate animals. Also, testing for indicator species can take several days, by which time the extent of the pollution might have changed.
Sogin's team has successfully used sequencing to detect a microbial profile of human feces in water samples, but "in the process of doing this, we have come to realize that if we want to trace these pollution zones that occur in environmental waters, we have to sequence much more deeply than we have in the past," he told In Sequence. "Even with the advent of massively parallel sequencing platforms, that has not been very tractable to do" because of the cost and labor involved, both limiting factors in running his lab's 454 machine.
Ion Torrent "gives us an instrument that we anticipate will have a much faster turnaround and have greater simplicity of operation," he said. Also, the cost of sequencing on the new platform will be "at least an order of magnitude less than what we're spending [on 454]," making it possible to increase the sampling density 10-fold and yet "have a very affordable experiment."
He said the 100-base read length "is good enough for us right now," though "the promise of longer reads is going to make this even more powerful."
Judging by the description of Ion Torrent's technology in its patents — the platform senses hydrogen ions released during the sequencing reaction — he expects the data accuracy to be higher than that of the 454 platform, though "the jury is out on that. We'll have to see what happens when it gets in our hands."
Likewise, the sample and library prep for the Ion Torrent platform — a "very intensive process" for the 454 platform — is also still unknown, he said.
Eventually, he hopes the technology could be used outside of a lab. "I have this fantasy that this technology can be reduced one day to a field-deployable instrument," he said. "That is going to really change things enormously."