About half of core facilities in the Northeastern US have already acquired a next-generation sequencer, or are planning to do so in the near future, according to an informal survey at a recent meeting of core directors.
While those who have bought an instrument are grappling with IT and bioinformatics challenges, others are still debating what next-generation instrument will be the winning bet for their institution, according to participants of the Northeast Regional Life Sciences Core Directors meeting at Cornell University earlier this month, which attracted more than 130 core directors from almost 60 institutions in 18 US states as well as Quebec.
Fueling the debate about what platform to buy is the variety of instruments on the market today. “Each platform has its particular strengths and weaknesses, so it is not such a cut and dry decision to make,” Bob Steen, director of the biopolymers facility in the department of genetics at Harvard Medical School, told In Sequence by e-mail last week.
Steen, who attended a DNA sequencing panel discussion workshop at the meeting, added that because of the high acquisition costs, core facilities usually only get a single shot at getting the right machine for their institution. “No one wants to pay a half-million dollars and end up with the next Betamax of next-generation sequencers,” he said.
But many core facilities have apparently already placed their bets. About half of the approximately 30 directors in attendance at the workshop raised their hands when asked who had already installed a next-generation sequencer, or was thinking of doing so in the near future, according to a participant.
George Grills, the meeting organizer and director of operations of core facilities at Cornell’s life sciences core laboratories center, said he and others were surprised to see so many labs bringing in new sequencing technologies so soon. “It seems like there will be many adopters of next-generation sequencing platforms in 2008,” said Steen. “[This will give] us more peers to discuss various technical challenges with and to ultimately improve the quality of service that we all provide to our end users.”
Last week, Steen’s facility installed an Illumina Genome Analyzer, and he said he hopes to offer full services on the instrument starting in January. When he and his colleagues were looking at available platforms in May, they evaluated 454’s, Illumina’s, and ABI’s instruments, based on a long list of criteria, including cost, ease of use, versatility, data quality, support, user experience, and availability of published data.
If demand for Illumina’s sequencing is great at his facility, his group will consider purchasing a second instrument, he said, re-evaluating all three platforms, and reviewing the instrument Danaher Motion has been developing in collaboration with George Church’s group (see In Sequence 10/16/2007).
According to Lee Ann Schein, director of the Robert Wood Johnson Medical School DNA core facility at the University of Medicine and Dentistry in New Jersey, “the technology is still so new that I would like to wait a few months until all of the machines are in use by cores.”
Her facility is interested in acquiring a next-generation sequencer, but “the price makes it prohibitive at this time,” she told In Sequence by e-mail. By absorbing her peers’ experiences, “I can learn the real-life pros and cons about each platform.”
Besides the initial cost, the greatest concerns of core directors are “the expense and the amount of time [and] work that prepping each sample takes,” Schein added.
Next-gen sequencing technology “is still so new that I would like to wait a few months until all of the machines are in use by cores.”
Jan Kieleczawa, laboratory manager of the DNA sequencing group at Wyeth Research, agreed that instrument cost, especially in relation to demand, is an important concern.
“For example, if you charge $15,000 for [a] 454 run, it will take you 40 weeks to recoup [the] costs of the machine,” he told In Sequence by e-mail. Many small core DNA sequencing labs may not have the need for a next-generation sequencer, he said.
His own facility is interested in acquiring a new sequencer, “most likely” a 454 or an ABI SOLiD, although he will give Danaher’s new instrument a look because of its low price of $130,000.
Among core facilities that already own next-generation sequencers, data handling appears to be the greatest challenge. “The biggest concern of our sequencing core director at the moment is how to manage (annotate, store, retrieve, etc.) all of the data coming off the instrument,” Chris Schad, a program manager for seven core genomics-related laboratories at Mayo Clinic, told In Sequence by e-mail. Mayo’s sequencing lab just took delivery of an Illumina Genome Analyzer, he said.
Grills echoed the need for “significant informatics support.” His center has had an Illumina sequencer for several months and recently returned a 454 sequencer after a demo period. The center is currently exploring ways to bring back the 454 instrument.
“We deal with a lot of non-typical plant and animal model systems that have no reference sequence, so it’s extremely important to have the [long read] capabilities that the 454 confers,” he said.
One topic of discussion, he said, was the question of how long to store the raw sequencing data. “I think everybody agreed that the current paradigm is really, ‘you save all raw data.’ That might change over time, but right now, a core would typically have at least some kind of backup of data.”
The main reasons are that samples may be irreplaceable, or users might want to take a second look at their data with new algorithms. “It reminds me very much of the early days of microarrays ... where folks might want to reanalyze their raw data to get more information out of it,” said Grills, whose center decided to back up all of its next-gen sequencing data for the first year with a plan to reevaluate its storage strategy at that time.
Also, because the technology is so new, core facilities have to help develop applications and analysis tools. “There are really very few standard protocols” for new sequencing applications like RNA profiling or ChIP sequencing, Grills said, and “the manufacturers provide usable but not optimal tools [for informatics analysis], so we have to really compare and help develop new tools.”
In addition, core facilities have to put up with “frequent upgrades to hardware and software,” he said, making standardized operating protocols and quality metrics difficult to achieve.
Another challenge is the need to schedule full runs and collect appropriate samples for them, he said, in order to run the instruments cost-efficiently.
Core facilities also have to decide in what form to return the data to their customers. If they analyze data for them, they need to be aware that the large datasets take “a lot of time to analyze, even with the largest clusters,” Grills said.
“Inevitably, the next-gen sequencing community will evolve the necessary tools to make this process more routine and manageable ... but in the meantime, core labs will need to be prepared with expertise to help design experiments and analyze data for many of their clients,” Steen said.