SAN FRANCISCO, Aug. 12 - Applied Biosystems' next-generation sequencer, scheduled for commercial introduction on Wednesday, is in the hands of early-access partners. So far they're impressed. The next challenge for ABI is to sell the tools in a receding sequencing market.
Funding for sequencing projects over the next several years is expected to stagnate. Plus, the number of ABI 3700s sold over the last few years has slowed as the drive to sequence the human genome has receded, said Omead Ostadan, ABI's product manager for DNA-platforms marketing.
Taken together, these two facts have hurt ABI's overall revenue growth over the past two years. In fiscal 2000 ABI reported a 24 percent increase in annual revenue growth. In 2001 that number fell to 17 percent, and in fiscal 2002, which ended in June, ABI reported a revenue decline of one percent, according to James Reddoch, an analyst at Banc of America Securities.
"That's the sequencer market talking," remarked an analyst who covers the Applera company.
Sales call ...
Hoping to boost sales of its sequencers, ABI has set its sights on hawking the 3100, which starts at about $80,000, to small- and medium-sized markets like core labs and biotechs. Meantime, the company will try to encourage 3700 users to trade up to the 3730, said Ostadan. The 3700 goes for approximately $300,000, and the 3730 begins at $250,000 for a 48-capillary unit and $350,000 for the 3730xl 96-capillary system.
Labs visited by ABI salespeople will likely encounter a pitch that highlights the 3730's greater efficiency and cost-effectiveness. By cutting costs in half, Ostadan said, the 3730 will give researchers access to not-so-sexy genomes.
Initial feedback from early-access partners, for whom 30 3730s were delivered this spring, has been mostly positive.
"The operators like" the 3730, said Chad Nusbaum, co-director of genome sequencing and analysis at the Whitehead Institute/MIT Center for Genome Research. His group received five of the machines and is one month into a three-month testing period. "It's easy to run," he said. "The robotics, the scary heart of the thing, seem to run well."
The 3730s "definitely have some real advantages in terms of automating the loading, and the read lengths look good," added a person close to the 3730 early-access program at the US Department of Energy's Joint Genome Institute.
This person, who spoke on the condition of anonymity, said JGI bought five of the 96 capillary-tube 3730s, but was unable to say what the institute paid.
Other early-access partners include the Sanger Centre, Washington University, and the Houston-based contract-sequencing company SeqWright, according to a number of people close to ABI's early-access program. Baylor University is currently negotiating with ABI to become an early-access partner, said an ABI spokesperson.
Nusbaum said ABI's claim that the use of reagent is lower with the 3730 than with the 3700 has so far been borne out, and that the increased automation should lead to lower labor costs. He also said that Whitehead is getting 15 to 20 percent more useable data from the 3730 versus the 3700, though he called this result "very preliminary" and added that the 3730 software makes "getting data on and off the machine a pain."
... but who's buying?
Will positive reviews translate into sales?
"I am not interested in adding capacity," Nusbaum conceded. "If we were to do a straight swap [with the 3700] we would do fewer [3730 purchases]." Nusbaum estimated that because of the higher efficiency of the 3730, the Whitehead Institute, which currently has approximately 168 3700s, would need to buy 10 to 20 percent fewer 3730s as direct replacements.
While the issue for Whitehead is going from one ABI machine to another, JGI, which owns approximately 21 Amersham Bioscience's MegaBACE 4000s, also needs to consider the impact of running sequencers from more than one company.
"The complication for us is not just how good is the ABI system, but we'd have to be a mixed house," said the JGI source.
Because of differences in loading capacity and throughput, this person estimated that the output of approximately 20 MegaBACEs would equal about 30 to 40 3730s.
At just a few weeks into JGI's testing of the 3730, the source said "the jury is still out" on a quality assessment that might lead to a purchasing decision.
Geoffrey Duyk, chief scientific officer of Exelixis, was more blunt. The 3730s "seem to be improvements in the quality of the data coming off the machine, but they're incremental [improvements]," he said. "It's like color TV five years ago versus now. Subtle differences."
While they may be subtle, the improvements are compelling enough for Exelixis to consider eventually replacing its 3700s with 3730s, said Duyk. And cost savings, especially at a time when sequencing funding continues to flatten, carries a lot of weight.
"Cost of sequencing is the crux of the issue," emphasized Nusbaum. "As long as cost continues to decrease, at least for several years there will be very big demand for data. People are clamoring for genomes."
To go to a bigger market, however, may require significantly reduced sequencing cost per base. Current technology "doesn't get you the thousand-fold decrease in cost that people want to see," said Duyk. "I think fundamentally you have to go to different technologies.
"I don't think we know what the technology looks like that will get to the next-generation sequencing," he added. "It's clearly not the technology that we have today."