By Julia Karow
Pacific Biosciences this week disclosed the first 10 early-access customers for its single-molecule real-time sequencer, which include genome centers and other academic institutes in North America as well as PacBio investor Monsanto.
The customers, who have ordered the $695,000 instrument at full list price, expect to receive their instruments this summer, with shipments expected to start in the second quarter, according to the company. The list price includes the main instrument console as well as a blade server for primary data analysis.
The first 10 early-access customers are Baylor College of Medicine, the Broad Institute, Cold Spring Harbor Laboratory, the US Department of Energy Joint Genome Institute, the Genome Center at Washington University, Monsanto, the National Cancer Institute/SAIC-Frederick, the National Center for Genome Resources, the Ontario Institute for Cancer Research, and Stanford University.
Notably absent from the list is Cornell University, where much of the basic technology for PacBio's platform was invented, but George Grills, director of operations of core facilities at Cornell's life sciences core laboratories center, told In Sequence that he hopes his institution will be among the second wave of early-access customers.
Later this year, PacBio plans additional early-access programs for customers in Europe and Asia, with the goal of making version 1 of the system, dubbed V1, widely available before the end of the year.
At the Advances in Genome Biology and Technology conference later this week, PacBio officials plan to reveal more details about the SMRT sequencer and its performance. In addition, scientific collaborators Elaine Mardis from the Genome Center at WashU and Joseph Puglisi from Stanford University School of Medicine are scheduled to present results from experiments conducted on PacBio's platform, including cancer genome sequencing and mutation detection and single-molecule analysis of protein translation.
The company's first commercial sequencer will process arrays with 80,000 zero-mode waveguides, tiny reaction chambers in which single polymerase enzymes synthesize DNA (see In Sequence 11/17/2009), while its prototype had 3,000 ZMWs per chip.
The average read length will be in the range of 1,000 to 1,200 bases, PacBio CEO Hugh Martin told In Sequence last week, and about 5 percent of reads will reach on the order of 3,000 to 5,000 bases in longer runs. Other performance specs, he said, "are still changing fairly dramatically as we continue to work on the enzyme in particular."
In the meantime, the company is working on a high-throughput "V2" implementation of its technology that will integrate sequencing reactions, optical detection, and signal processing in a single chip. That system is scheduled for release in 2014 (see other article, this issue).
While some of the early-access customers were collaborating with Pacific Biosciences prior to their instrument order, others were not. Several said the system appears to have distinct advantages over other sequencers, and expressed confidence about its performance and future improvements.
"We are extremely excited about the potential of this instrument to do things other technologies currently can't, such as producing very long reads, producing highly accurate reads by sequencing the same fragment several times in the same read, and generating data with very short run times," said Chad Nusbaum, co-director of the genome sequencing and analysis program at the Broad Institute.
[ pagebreak ]
"We like the quick turnaround times, in comparison to the other instruments we currently run," agreed Michael Smith, director of the genetics and genomics group at SAIC-Frederick. Smith oversees two core laboratories for the National Cancer Institute, which currently run four Illumina Genome Analyzer sequencers and two 454 GS FLX instruments.
PacBio said previously that an entire experiment, including library preparation, will take less than 12 hours on its platform. According to Smith, library prep alone on the Illumina system takes a day for genomic DNA, and considerably longer for other library protocols.
Fast turnaround, he said, will be especially useful for cancer patient characterization studies that he and his colleagues are planning. "If you do that kind of work, you cannot afford to wait for two weeks," he said.
Another advantage of PacBio's platform, he said, is the relatively low cost per run, so researchers can do more "on-the-fly" quality control of samples to make sure the correct amount of DNA is loaded on the instrument. He said that samples submitted by investigators to the core labs vary in quality, "so you can envision a run on a machine serving a quality control purpose before you go and do, perhaps, multiple runs to get a full dataset." Last year, PacBio said that consumables and reagents for a single run will cost $99.
Smith said he expects the throughput and read length of the system to improve considerably in the future, and looks forward to hearing more from the company at the AGBT meeting. "If they are able to get out to several thousand base pairs as a minimum, it puts the machine in a class of its own," he said.
Asked whether it is worth paying full price for a new technology that has not been tested by others, he said that "Obviously, one takes risks. But we put our money on the table."
According to Dick McCombie, a professor at Cold Spring Harbor Laboratory, being an early adopter of a new sequencing technology has paid off for his institute in the past.
He and his colleagues received one of the first Solexa (now Illumina) sequencers at the end of 2006, paying full price. Although testing the platform and bringing it into production mode took time and effort, he said, it allowed the CSHL researchers to learn "a lot about how to most effectively change the way we do a lot of research institution-wide to take full advantage of having this disruptive technology."
He said he expects the same will happen with PacBio's technology, adding that he is already getting requests from graduate students wanting to adapt new research projects to the platform. "I think being in early helps catalyze that creativity," he said.
Cold Spring Harbor researchers are currently using both targeted and whole-genome sequencing to look for genetic and epigenetic variation in psychiatric disorders and cancer, and the PacBio instrument will improve those studies "in terms of increased speed and decreased cost in the long run," McCombie said. In addition, it will "likely allow us to look at structural changes with a lot better resolution than is currently possible."
Besides disease studies, he and his colleagues plan to apply the PacBio instrument to "obtain genomic information from plant genomes that have not been sequenced previously."
"Genetics has really been transformed into genomics over the past three years, and we intend to keep pushing those capabilities with the PacBio instrument," McCombie said. "I'm expecting it to be both very challenging and a lot of fun." Although his institute has not been collaborating with PacBio, he said he has "a lot of confidence" in the company, based on people he knows there.
Researchers at the Ontario Institute for Cancer Research plan to apply their PacBio sequencer to individualized medicine, John McPherson, director of cancer genomics at OICR, told In Sequence. "The speed and read length of the instrument are attractive for this application," he said.
Although the initial throughput of the instrument will limit the amount of "gene space" that can be resequenced in parallel, he said, "we anticipate rapid improvements as the instrument matures." In the meantime, he and his colleagues are planning several oncolytic virus sequencing projects "that the platform is ideally suited for at its current scale."