Affymetrix this week announced that it has begun limited shipment of its single-chip SNP 5.0 Array on schedule, a consolidated version of the dual-chip 500K array set the firm launched in the third quarter of 2005.
The company also delayed by three months the launch of its two-chip, 1-million-SNP product, which was originally scheduled to launch in the first quarter of this year. Affy is betting that the so-called SNP 6.0 array, now due to appear in the first half of this year, will raise the stakes in the competitive whole-genome genotyping market (see BAN 7/25/2006).
Separately, Affy said it expects fourth-quarter revenue of $100 million, a decline of 11.5 percent over the same period last year.
Affymetrix did not respond to an e-mail seeking comment about its decision to push back the launch date for the 6.0 set. However, CEO Stephen Fodor told investors at the JP Morgan Annual Healthcare Conference in San Francisco this week that the company is currently selecting content for the product from a master list of 1.6 million SNPs.
Fodor also discussed the company’s upcoming fourth-quarter financial report, which is expected to show $100 million in revenue compared with $111.5 million in the fourth quarter of 2005. Despite being an 11.5-percent decline, the revenue estimate beat Wall Street analysts’ expectations, which projected the company would post $99.5 million for the period.
Fodor said Affy experienced “solid performance” in Q4 and partially attributed this to its decision to cut prices for its 500K Array Set by half to $250 last July (see BAN 7/25/2006). ” When we dropped them down to $250, we got a 90-percent increase in orders,” Fodor said at the JP Morgan conference. “In Q4, we saw an incremental increase in orders beyond that.”
“Suffice to say, this has led us to push on the genetic power, the content, and cost in this field,” he said.
Central to that strategy has been the launch of the new SNP 5.0 array, which the firm said features SNPs from the original two-chip 500K array set, as well as 500,000 additional probes that can measure other genetic differences, such as copy number variation. The pricing for the single-chip set will remain at the same level as the two-chip set.
“This chip has 500,000 SNP assays and in addition has another 500,000 copy number measurements,” Fodor said of the 5.0. “In essence, what you can do with this chip is look at the 500,000 human variations and do a measurement of 1 million measurements on copy number across the genome.”
Affy said in a statement that it developed the SNP 5.0 Array with the Broad Institute of Harvard and the Massachusetts Institute of Technology to “better identify and understand the genetic variations associated with complex diseases such as autism, autoimmunity, bipolar disease, cancer, diabetes and heart disease.”
David Altshuler, director of the Program in Medical and Population Genetics at the Broad, said in the statement that the institute’s investigators “want simultaneously to test both SNPs and copy number variants, all in the same experiment” in order to “understand the causes of complex genetic diseases.”
Customers Express Interest in New Chips
Florian Wagner, head of the Affymetrix Service at the Deutsches Ressourcenzentrum für Genomforschung (RZPD) lab in Berlin, said his facility is interested in switching from the two-chip 500K set to the single-chip 5.0 array and then to the 1 million SNP 6.0 product.
“RZPD offered the 500K SNP genotyping service for some time, but we stopped this about three months ago, because we experienced too many problems,” Wagner wrote BioArray News in an e-mail last week. “Some chip lots were of minor quality, the assay itself seemed to be not really stable, and in many cases we were not really satisfied with the call rates, even if they exceeded 93 percent,” he wrote.
“We are now waiting for the 500K single chip, and of course the 1 million chip set, and we hope that the assay itself as well as the chip production will be more stable, so that we get consistently high call rates,” Wagner explained.
In addition, Wagner said that the 5.0 product “will be very interesting for very high resolution of copy number changes and loss of heterozygosity” studies and that he expects that “clinicians doing basic research [in areas like] cancer and mental retardation will be among the first to use the new chips.”
During the conference this week, Fodor acknowledged that customers like Wagner had had difficulties with launching the dual-chip 500K set, and said that he expects the transition to the new chips to go more smoothly because the firm has done extensive beta testing and its accumulated know-how is much higher.
Rob Hall, who manages the Center for Array Technologies at the University of Washington in Seattle, told BioArray News this week that the CAT will definitely adopt the new technology but that its clients will most likely wait until peer reviewed data from the SNP 6.0 array becomes available.
“We will definitely train our genotyping staff on the 1 million SNP protocols and offer the service to our customers,” Hall wrote via an e-mail. “I believe our ongoing projects will continue to use the 500K set until there is sufficient data available from peer review to indicate that the 1 million SNP product is reliable and appropriate to researchers' needs from both a biologically relevant and cost-effective viewpoint,” he wrote.
“We are now waiting for the 500K single chip, and of course the 1 million chip set, and we hope that the assay itself as well as the chip production will be more stable, so that we get consistently high call rates.”
However, Hall said that there will “certainly be some immediate interest” in the new arrays. “Being able to assay 1 million SNPs on a single chip is really a remarkable achievement, and opens up a lot of possibilities for increasing the resolution and statistical power of large data set analysis in both linkage and population genetics studies,” he wrote.
“I think the degree to which the interest in the one million SNP product is maintained and increased will be dependent upon the research published using the product, ease of use, pricing, and whether or not competing companies are able to develop comparable or even superior genotyping platforms and applications in the near future,” he added.
The big players are already beginning to compete on price. In July, Illumina CEO Jay Flatley said that Affy’s decision to cut prices and issue new products would have “no direct impact” on Illumina’s pricing going forward (see BAN 7/25/2006).
However, a few weeks later, Illumina cut its prices. As reported by BioArray News sister publication Pharmacogenomics Reporter in October, Illumina dropped the list prices and reduced the discounts for its Infinium whole-genome genotyping kits in September. Flatley said at the time that the change “was probably precipitated by Affy’s change, but it didn’t cause it.”
According to Dietrich Stephan, director and senior investigator of the Translational Genomics Institute’s neurogenomics division, the genotyping market could very well come down to density in the future.
“This whole game comes down to three things: price point, throughput, and density. And accuracy is a function of density. If you double the density on a chip, then the chip will be more accurate than a competitor’s chip,” Stephan told BioArray News last week.
“Ultimately you want the highest density in order to get the highest accuracy. It is a numbers game,” he said. “I think that when these tools start getting to the 1 million SNP mark, then researchers will start having enough statistical power to do studies as if they were resequencing the entire genome.”
He added that even higher density chips are likely to come out in future years. “For some of the younger [genomic] populations [1 million SNPs] is probably sufficient, but for more ancient populations we are going to need three million to four million SNPs to have significant power to do disease detection studies,” he said.
“I think it is common knowledge that Affymetrix will continue to shrink feature size to put more content on the arrays. I can easily see a 4 million-SNP product in the future,” Stephan said.
With additional reporting from Edward Winnick