The market for second-generation sequencing technologies is poised to more than triple over the next five years and will be dominated by research applications, while the technical performance of current second-generation platforms will improve severalfold during that time, according to assessments by the three dominant market players.
At the Leerink Swann Emerging Products and Applications in Life Science Tools roundtable conference in New York this month, Applied Biosystems, Illumina, and 454 Life Sciences gave an update on their views of the sequencing market and the performance potential of their respective platforms.
Illumina and ABI are currently competing for market share with their short-read technologies, while 454 is focusing on applications for long reads, including most of the traditional Sanger market.
“As more studies get published, you are going to see pretty quick consolidation” among the current market players, Isaac Ro, vice president for equity research at Leerink Swann, told In Sequence during the conference. “One or two are going to get the lion’s share of the market, and these other guys are going to find their niches.”
The market for second-generation sequencing will more than triple over the next five years from $200 million in the current fiscal year to $645 million in fiscal year 2013, according to Shaf Yousaf, president of ABI’s MCB Genomic Analysis division, who spoke at the conference.
The estimate, which focuses on current second-generation platforms and does not include third-generation technologies that might enter the market over the five-year timeframe, is based on a new study that consulting firm L.E.K. recently completed for ABI.
In terms of sequencing applications, more than half of the market in 2013 will consist of “gene structure” projects that investigate the structure of genomes, both of healthy and sick individuals. The second-largest application will be mRNA gene-expression studies, followed by “gene regulation” studies that include, for example, the analysis of non-coding RNAs.
The market in 2013, which is estimated to be worth around $650 million, will be dominated by short-read sequencing platforms, which will capture more than $500 million with the remaining $150 million or so going to long-read platforms, according to the study. “Short-read technologies will dominate the market for the foreseeable future,” according to Yousaf.
Academic customers will remain the “key driver” for the market, he said, and are expected to spend $400 million on second-generation sequencing by 2013.
As workflows for operating second-generation instruments become easier, and data analysis becomes more routine and supported by standardized tools, the platforms “will be adopted by typical molecular biology labs, not just the genome centers,” Yousaf said, and will be used “across academia very widely.”
Other customer types, including clinical researchers, biopharmaceutical customers, and agbio customers, will also increase their use of next-gen sequencing, “but the growth will be coming largely from the academic side,” he said.
“You really need the diagnostics market to be a credible opportunity” for second-generation sequencers.
ABI’s analysis apparently does not include the market for “traditional” Sanger-based sequencing, which Yousaf did not address in his presentation. A year ago at the same conference, the company said it expects the sequencing market for all technologies to grow from $1 billion in 2007 to $1.4 billion in 2011, with next-generation sequencing growing from $135 million to $450 million over that period (see In Sequence 8/14/2007).
The company estimated the current market size for capillary electrophoresis-based Sanger sequencing to be approximately $900 million last year. It is this market segment that 454 Life Sciences has set its eyes on with its Genome Sequencer FLX technology, which provides longer reads than ABI’s and Illumina’s platforms.
According to Michael Egholm, 454’s vice president of research and development, the company’s goal is to take over most of today’s Sanger sequencing market and grow it further. He estimated that 454 can command between 70 percent and 80 percent of that market, which he said is around $700 million today.
454, which was acquired by Roche last year and no longer discloses revenues or numbers of instrument placements, has already made headway in its quest to take over Sanger sequencing, according to Egholm. He said that 454’s business doubled in the second quarter of 2008 compared to the same quarter in 2007.
He also disclosed the number of 454 platforms placed at large genome centers: Baylor’s Human Genome Sequencing Center and the Broad Institute each have 10 instruments and Washington University’s Genome Sequencing Center has eight machines.
Illumina CEO Jay Flatley gave a more quantitative assessment of the market for new sequencing technologies. In the research market, he said, “every lab is a candidate target” for an Illumina Genome Analyzer, as second-gen sequencing is becoming a requisite for genomic research.
“In order to be a competitive lab, you need to have access to this technology,” he said, adding that the clinical diagnostics market presents opportunities “mostly in areas like cancer,” where sequencing could play a key diagnostic role in the future.
Yousaf agreed. “I have spoken to some academic researchers that believe that eventually, every cancer will be sequenced, and in fact, the way of characterizing a particular cancer and designing a therapeutic around it will involve sequencing that cancer,” he said.
Looking at the consumer genomics market, Flatley said it will take at least three to five years before sequencing will enter that space, as array-based genotyping technologies are still “a lot less expensive” than sequencing. For now, consumer sequencing will cater to the “rich and famous market,” he said.
According to Leerink Swann’s Ro, the long-term promise of next-gen sequencing lies in markets other than research. “You really need the diagnostics market to be a credible opportunity,” he said. “That’s something that you can see, [but] it’s not there yet.”
Illumina, ABI, and 454 also gave a glimpse of future performance specs of their respective platforms.
Over the next few months, 454 plans to roll out its new Titanium chemistry and plates, which will allow its GS FLX to produce up to 500 megabases of data per run with an average read length of about 500 bases. By including an enzyme in the new sequencing kit that used to be sold separately, reagent costs will drop by 10 percent to 15 percent, according to Egholm.
Researchers at Baylor have already used the updated technology to assemble the 180-megabase Drosophila genome de novo. Next in line for 454 are a de novo assembly of a 1-gigabase genome later this year, and the company hopes “to do human de novo sequencing within a year or so,” according to Egholm.
Within the last 18 months, the company has increased the throughput of its platform by fivefold, and “there is nothing to prevent us from doing it again,” he said.
According to Flatley, Illumina expects customers to soon obtain 5 to 7 gigabases of data per run on its Genome Analyzer II, and some, depending on certain unspecified factors, “considerably more.” By the end of the year, customers will be able to obtain 20 gigabases per year, he said.
However, overall throughput is going to be “less and less of a factor” in the future, he predicted. What is going to become increasingly important, as second-generation sequencers gain traction with smaller labs, is ease of use and data quality.
Higher-quality data, Flatley pointed out, also lowers the fold coverage, or throughput, needed in a given sequencing project. Increasing the quality score from Q20 to Q30, for example, could enable researchers to reduce the necessary coverage of a genome from 30X to 20X.
Flatley said Illumina is also working on increasing its system’s read length from the current 50 base pairs to 75 base pairs — both for fragment reads and for paired-end reads — by the end of the year, providing several insert sizes for paired-end sequencing.
“From 75 [bases], we see a very clear runway to 100, and there is nothing we have seen yet that prevents us from going beyond 100,” he said. “For some applications, this is going to be important, for others, not important.”
With a read length of 100 base pairs, said Flatley, “most people believe that you can do de novo sequencing,” though he acknowledged that there will be “some applications that will require longer read lengths.”
“We think once we get to 100 bases, we will have access to most applications,” he said.
According to ABI’s Yousaf, some SOLiD customers currently obtain 20 gigabases of data per run on the platform, and “internally, we are pushing the forefront of 30-plus gigabases” per run.
The company’s “near-term goal” is to get to 50 gigabases per run, “and we believe that with further changes in protocol and chemistry, we can achieve this,” he said.
That throughput, he said, enters the range of what so-called third-generation sequencing technologies promise to deliver, and “we believe that we are pushing this with our [existing] next-gen technology,” he said.
Also, “there are aspects [of the technology] that we haven’t even looked at,” he said, including bead size and dyes. “Improving those could take this even further,” he said.
Internally, ABI currently obtains read lengths of 50 bases in paired-end sequencing and 75 bases in fragment-based sequencing.
In order to allow users to run multiple samples per run, the company also recently introduced barcodes for multiplexing. Each run can now accommodate up to 256 samples. Two of its customers, which were not disclosed, are “getting excellent results with multiplexing in small RNA research,” he said.