Applied Biosystems is starting to drum up business for its SOLiD next-generation sequencer, which it plans to begin rolling out to early-access customers this summer.
Last week, the company invited about 50 scientists from around the world, most of them potential SOLiD customers, for a three-day meeting at its Foster City, Calif., headquarters.
ABI has reason to get the ball rolling: It will likely be the third vendor to bring a next-generation sequencing system to market, trailing 454 Life Sciences, which sold its first Genome Sequencer 20 two years ago and had placed more than 60 by the end of 2006, and Illumina, which shipped 13 Genome Analyzers to early-access customers in 2006 and had received at least 40 orders for its system as of early February.
ABI, which has dominated the market for capillary electrophoresis sequencing for years, is optimistic that it can stand up to the first movers.
“We have confidence in our system,” Haley Fiske, ABI’s director of product management for high-throughput discovery systems, told In Sequence last week in a phone interview after the meeting. “We believe our system will produce more data at better cost-benefit to the customer” than competitors, he said.
The cost per “mappable” gigabase — that is, good quality sequence data — will be around $3,000, Fiske said, including sequencing reagents and library construction. The price could be higher, though, if users run more than one sample per slide, he cautioned.
454 Life Sciences has recently cited a cost of $8,000 for reagents per 100 megabase run on its GS FLX, and Solexa has said in the past that flow cells and reagents for its Genome Analyzer would cost between $3,000 and $5,000 per run, which can generate up to one gigabase of data at the moment.
Initially, ABI’s SOLiD sequencer will yield between 20 million and 40 million mappable reads per run — translating into 1 to 2 gigabases when using a paired-end library — and by next year could yield 80 million reads per run, distributed over two slides. But “we have strategies to take it beyond that — double or triple that,” Fiske said.
To make its point, ABI showcased its SOLiD technology to invited researchers from the UK, Germany, France, Belgium, Spain, Sweden, Switzerland, Iceland, Australia, Japan, China, and Korea, who attended last week’s meeting on the first two days, and from North America, who came to Foster City on the third day. Attendees received a tour of the company’s labs, including a view of early prototypes of the instrument, and listened to talks from high-level company representatives as well as outside collaborators.
The company chose the meeting participants, many of whom already own other ABI instruments, because “a lot of them have very interesting projects that they would like to see done on next-gen systems,” Quynh Doan, ABI’s product line manager for high-throughput systems, told In Sequence last week. ABI’s sales force suggested who to invite, she said.
In addition, ABI specifically targeted those researchers and institutions who did not attend the recent Advances in Genome Biology and Technology conference in Marco Island, Fla., where ABI presented the SOLiD technology, and “where a lot of genome centers and lots of top US customers were [present], but not necessarily from around the world,” Doan said.
The company is currently collaborating with more than 10 research institutions on projects that involve SOLiD sequencing, according to an ABI spokeswoman. Five of these collaborators presented at the meeting, covering a variety of applications.
Darren Platt, informatics department head at the Department of Energy’s Joint Genome Institute, presented a preliminary assembly of the Streptococcus suis genome using a combination of SOLiD data, which ABI had only given to him a few weeks before the meeting, and capillary electrophoresis sequencing data from the Wellcome Trust Sanger Institute in England.
He co-assembled the data using Forge, a genome assembler he developed. The scaffolds in the joint assembly were considerably longer than those in an assembly of Sanger data only, he said.
“The SOLiD data definitely helps improve the assembly,” Platt told In Sequence by e-mail this week. One of the major challenges was the sheer volume of the data, he said — 23 million reads in total — which took about two days on 11 computers in a cluster to assemble. Following some further optimization to the assembly process, he now plans an assembly from SOLiD data only.
Sean Grimmond, a senior research fellow at the University of Queensland in Australia and head of the UQ Institute of Molecular Bioscience’s expression genomics laboratory, also talked about “the back-end bioinformatics” of SOLiD sequencing, including how to deal with several gigabases of sequencing data; models of transcriptional complexity, or isoforms, to determine how deep to sequence mRNA in order to monitor gene activity, alternative splicing, and promoter usage; library design; and quality control of samples.
He has been collaborating with ABI for about six months on a shotgun transcriptome sequencing project to study embryonic stem-cell differentiation.
“We have just completed the sequencing and the computers are currently crunching data,” he told In Sequence in an e-mail message last week. His institute is also interested in using the SOLiD technology for chromatin immunoprecipitation sequencing, resequencing bacterial genomes, and, later on, resequencing clinical samples.
“We have spent a long time reviewing all the other technologies,” Grimmond wrote, in particular 454’s. “The decision to engage AB’s SOLiD sequencing over competing platforms came down to low operating costs, the platform’s scope to rapidly increase the volume of sequence obtained per run, and the level of support provided” by ABI.
But the current projects are “still definitely work[s] in progress,” he wrote, and he is benchmarking the sequencing results against the “gold-standard array platforms” – Illumina’s BeadArray system and Affymetrix’s exon arrays — using the same samples.
Stephen Kingsmore, president of the National Center for Genome Resources, talked about his experience with medical resequencing, focusing on sequencing brain samples from schizophrenia patients. He addressed the methods and approaches involved, and the software needed to handle next-generation sequence data.
While his lab has mostly generated data on 454’s platform so far, it is now also obtaining SOLiD data from ABI. “The data has been generated, but we did not have time to pipeline it for the meeting,” Kingsmore told In Sequence this week.
The two remaining collaborative projects had already been presented at last month’s AGBT meeting: Arend Sidow from the departments of pathology and genetics at Stanford University Medical School talked about using SOLiD sequencing to position nucleosomes in the C. elegans genome, and Paul Richardson from the JGI showed genome resequencing data for Pichia stipitis from a collaboration with both ABI and Agencourt Bioscience (see In Sequence 03/07/2007). That project is actually a broader comparison between ABI’s SOLiD platform, Illumina’s Genetic Analyzer, and 454’s Genome Sequencer.
Participants in the Foster City meeting said the event was useful to them and answered their questions about the technology. “I was very impressed how open they are,” Kingsmore said, referring to ABI. “They are really bending over backwards to tell people what they are doing, to solicit feedback, and to field questions.”
“This is a brand-new market, and [ABI is] very keen to get a lot of market feedback.”
“What they offer, and what they showed us, and what collaborators talked about was very interesting,” agreed Markus Schilhabel from the University of Kiel in Germany. “Even though, to be honest, coming from genome sequencing, I am always keen on having long reads.”
Schilhabel is currently setting up a new sequencing lab at the university’s Institute of Clinical Molecular Biology that will house a SOLiD system (see In Sequence’s sister publication, GenomeWeb News, 12/19/2006).
His lab plans to use the instrument initially mostly for transcriptome analyses, as well as for resequencing targeted areas of the human genome. Schilhabel sees the low cost per base as a main advantage of the SOLiD system, especially compared to 454’s GS FLX, which his laboratory will also acquire.
But 454’s technology still has the advantage of comparatively long reads. “For me, the SOLiD and the Solexa [instruments] are competitors, and the Roche [Genome Sequencer] is more a competitor to the normal capillary sequencers,” he said.
Kris Lambert, an associate professor in the department of crop sciences at the University of Illinois at Urbana-Champaign, said that he was especially impressed by how the SOLiD platform discovers SNPs. He said he came to the meeting because he is interested in using next-generation genotyping to analyze plant parasitic nematodes, which have been poorly characterized at the genomic level so far.
“Since I have been trying to find SNPs with 454 sequence data currently, I can see how the SOLiD system will be much better than 454 for this application,” he told In Sequence by e-mail last week. “Also, in my experience, the 454 sequence has many indel errors that can be annoying,” he wrote. “However, the longer sequence reads generated by 454 are very useful for EST projects and possibly for collecting a low level of coverage for de novo genome sequencing.”
After generating some preliminary data on the SOLiD platform, “I would then consider trying to raise the money for an instrument for our university,” he wrote.
For ABI, it seems, it will be key to get potential users hooked before the competition lures them first, especially customers that will not need more than one next-generation sequencer.
“I am sure…there are people they want to persuade to wait six months rather than buy one of the competitors’ boxes,” Kingsmore said. “But I also think that this is a brand-new market, and they are very keen to get a lot of market feedback...and that this process helps them a lot to [determine] what the applications are likely going to be.”