As Roche, Illumina, and Applied Biosystems spread the word about their new sequencing platforms, the diverse applications associated with these technologies are spurring some academic core facilities to buy an instrument of their own.
“I think the interest is tremendous,” said Kevin Knudtson, director of the DNA facility at the University of Iowa and organizer of a session on DNA sequencing at last week’s Association of Biomolecular Resource Facility’s annual meeting in Tampa, Fla.
A number of core facilities have already installed a 454 sequencer: the Interdisciplinary Center for Biotechnology Research at the University of Florida, Gainesville, has had a GS 20 since mid-2005 and recently upgraded to the GS FLX (see In Sequence 4/3/2007); the Biomedical Genomics Center at the University of Minnesota has been operating a GS 20 since last November and is planning to upgrade to a GS FLX this spring. Two weeks ago, Cornell University’s Life Sciences Core Laboratories Center had a 454 sequencer installed, and is expecting to receive an Illumina Genome Analyzer later this spring.
Other core labs, including Knudtson’s, are still deciding whether to buy a next-gen platform, and if so, which one. “A number of my colleagues specifically came to that meeting to learn more about these second-generation sequencers,” he said.
“More and more investigators at our [research] institutions are demanding that we maintain these [next generation] instruments [so they can] be competitive with their colleagues that have access to that [technology],” he said.
It seems that core facilities that as recently as a year ago thought they had no need for a high-throughput sequencer are now realizing that the new platforms offer more than just whole-genome sequencing.
“I think more and more folks are starting to realize the value and the capability of doing directed resequencing, as well as transcriptome-based studies, or small RNA profiling studies, or protein-nucleic acid association studies… on these kinds of platforms,” said George Grills, director of operations of core facilities and advanced technology assessment at Cornell University’s life sciences CLC. He gave a talk at a “well-attended” Roche/454 workshop during the ABRF meeting.
ABI and Illumina also presented their new technologies at the meeting. According to Bob Nutter, senior specialist for product applications at ABI, “core lab managers are very interested in new sequencing technology, but in general, are unsure how they would incorporate the technology — SOLiD or other — in their labs.”
He told In Sequence that many managers he talked to at the meeting do not know yet how such an instrument would fit into their business model, and that most are still looking for funding sources.
According to Elaine Mardis, co-director of the Genome Sequence Center at Washington University, “people are very excited about the potential for the technology, and rightfully so, but there is baggage that comes with it that isn’t insurmountable, [but] you just have to be sure that you are aware of it.”
Mardis gave a talk at the meeting about the GSC’s experience with 454’s and Illumina’s new sequencing platforms that included many practical considerations and cost estimates.
The first thing core facilities should consider when thinking about buying a next-gen sequencer is the applications they want to use it for, she told In Sequence.
For that reason, before ordering an instrument for his facility, Grills surveyed potential users at Cornell about their needs. “We identified about 30 investigator groups with more than 30 projects who want to use these platforms,” including agricultural, veterinary, and medical studies, he said.
Such an analysis might help not only decide whether to bring in a next-generation sequencer at all, but which one – Roche’s 454 Genome Sequencer, or Illumina’s Genetic Analyzer – or whether to wait for ABI’s SOLiD technology, which will be shipped to early-access customers this summer.
“There are different sweet spots for the 454 or the Illumina instrument,” Grills said. Also, “there is an interesting debate at the moment as to the relative merits of the SOLiD versus the Illumina instrument,” he said, which appear similar in terms of read length and amount of data per run, at least in their initial versions, although only the Illumina instrument is available now.
But it is only on the basis of real data – preferably from peer-reviewed publications — that users can make an informed choice. “There are a lot of good theoretical advantages of one or the other platform, but the proof of the pudding is going to be in the data that is actually published,” Grills said.
At the Biomedical Genomics Center of the University of Minnesota, the main application of the GS 20 is whole-genome sequencing, for example of pathogenic strains of bacteria. “This is the reason why we looked at the technology in the first place,” said Sushmita Singh, a research associate and scientific coordinator at the center.
Prior to acquiring the instrument, the cost of sequencing 5 million base pairs by Sanger technology at the center was almost $200,000. “Not everybody can afford that,” she said. On the new instrument, this price has come down to about $20,000, she said.
Consider Computational Needs
Users of the next-gen sequencers stress that people should not underestimate the computational needs, space requirements, and startup costs. According to ABI’s Nutter, many core labs do not have the bioinformatic infrastructure that is needed to support a high-throughput sequencer.
“I think more and more folks are starting to realize the value and the capability of doing directed resequencing, as well as transcriptome-based studies, or small RNA profiling studies, or protein-nucleic acid association studies … on these kinds of platforms.”
Mardis mentioned in her talk that one run on 454’s GS FLX generates about 30 gigabytes of raw data, and a single run on Illumina’s machine creates almost 650 gigabytes of image data.
While it might be tempting to throw the image data out after a while — since the cost of producing the data again might be cheaper than storing the images long-term — that might not always be a good idea. For precious clinical samples from patients who have since died, for example, “you will never have access to that again,” Mardis said.
It is also important to have fast data transfer, ideally gigabytes per second, and sufficient CPUs for processing the data. “These are just things that a lot of core facilities have not had to deal with before,” she said.
So is it even feasible to run a next-gen sequencer outside of a large genome center? “The answer to that is ‘yes,’ as long as you are taking care of the informatics and IT needs,” Grills said. His facility, for example, has access to a 400-node cluster to support its sequencing applications.
Singh’s center, meantime, is working with the University of Minnesota Supercomputing Institute, which has several bioinformaticists on staff.
Finally, potential users need to be aware of the cost – not only for the instrument itself, but also for ancillary equipment, data storage, and for running the instrument.
While some core facilities have turned to internal funding sources, including contributions from departments or institutional programs for research infrastructure, others are applying to shared instrumentation grants from the National Institutes of Health or the National Science Foundation.
“Right now, that is one of the challenges at our institution, coming up with the money upfront,” Knudtson said.