As part of an effort to provide its current customers with the latest rare variation content for genome-wide association studies, Illumina will roll out a supplementary BeadChip for its HumanOmni2.5-Quad DNA Analysis BeadChip alongside the launch of its HumanOmni5 BeadChip later this year.
Carsten Rosenow, Illumina's associate director of global market development for genotyping, told BioArray News this week that the San Diego vendor will launch an array containing more than 2.5 million SNPs when it rolls out the 5-million marker HumanOmni5.
The supplementary array will enable customers who have adopted the firm's recently launched HumanOmni2.5-Quad DNA Analysis BeadChip to access the new rare variant content provided on the HumanOmni5, which is scheduled to be launched later this year.
Since last year, Illumina has been releasing new products in its Omni series of arrays that contain rare SNPs and copy number variants generated by the 1000 Genomes Project and other sources.
The company said it has been able to keep up with new batches of 1000 Genomes Project data by filtering through it and designing new chips fitted with what it considers the most useful content.
Rather than make its customers wait for a single chip to power future rounds of GWAS, however, Illumina developed a "roadmap" designed to launch new chips together with supplements. In this way, customers who started projects with older arrays can keep pace with the latest content by running their samples on the supplementary chips.
The idea of the roadmap was to "enable any researcher who has started with any Omni product to scale up their projects all the way up to the five million SNPs we anticipate will come out of 1000 Genomes for SNPs down to a minor allele frequency of 1 percent," Rosenow said.
As an example, when Illumina began shipping its HumanOmni2.5-Quad earlier this month, it also began shipping its HumanOmni1S-8 BeadChip, which allows current or former users of Illumina's Omni1-Quad or OmniExpress BeadChips to access new content from the 1000 Genomes Project- without having to purchase the Omni2.5.
"Just like the 1S allows [users of the] OmniExpress or Omni1-Quad BeadChip to get close to the content of the 2.5-Quad, the 2.5S will bring any researcher from the Omni2.5 to Omni5," Rosenow said.
Adhering to the Illumina roadmap before the Omni5 launches will require customers to run more chips, but the company has developed a number of incentives, including price, to encourage adoption of the current Omni arrays and their supplemental chips.
According to Rosenow, "running the supplemental arrays will actually save you money," as opposed to running samples on the Omni5.
The company is charging $125 per array for the 1S chip, which contains eight separate arrays, and expects to charge $175 per array for the 2.5S supplementary chip, though Rosenow noted that prices in local currencies may vary.
He declined to discuss the price of the Omni2.5, but said that the combination of the 1S with the OmniExpress or Omni1-Quad will cost less than the Omni2.5 would on its own.
Another incentive that Illumina expects to encourage customers to perform GWAS on its chips is the availability of a multi-sample preparation kit that allows customers to prepare one sample for use on multiple chips over time. Rosenow said it was "always feasible" for Illumina to provide such a kit to its customers, but that once Illumina embarked on its roadmap, customers pointed out the challenge of doing a new DNA prep each time they adopted a new array.
"Any customer who wants to go on roadmap and use supplemental arrays only has to do DNA prep once," Rosenow said. "That can be used from beginning and then moved all the way up to the Omni5," he said. "This really simplifies the workflow for customers who want to go on the roadmap."
Allowing customers multiple entry points to running studies on this next generation of chips is a priority for Illumina, which is just now beginning to see its array business bounce back after a lull in the GWAS market last year.
CEO Jay Flatley told investors at Goldman Sachs Global Healthcare Conference earlier this month that Illumina has designed its roadmap to the Omni5 to give its customers more options when it comes to arrays (BAN 6/22/2010).
"We have created the ability for customers to start at different positions in this roadmap," Flatley said. "You could start with the Omni2.5 or the OmniExpress and then very economically get up to the Omni2.5 or the Omni5," he said.
According to Rosenow, Illumina settled on the endpoint of its roadmap early on, but believes that the content on its Omni2.5 chip validates its idea of how many SNPs would be required to offer customers access to the wealth of data being generating by 1000 Genomes Project. The project last week released data for its pilot phase, and the two-year production phase of the effort, which ultimately aims to analyze the genomes of as many as 2,500 people, will end in early 2011.
Rosenow said that the 1000 Genomes Project has currently discovered about 17 million SNPs, and that Illumina anticipated it may ultimately generate 35 million SNPs. Still, the company believes the use of tagSNPs will allow it to account for those 35 million SNPs by using around 5 million SNPs on one chip.
"These are numbers extrapolated from all findings in the 1000 Genomes Project based on the early identification of novel SNPs at minor allele frequencies," said Rosenow. "For this milestone, we have hit it right on with the [Omni2.5]," he said, noting that the company's most recent BeadChip interrogates around 80 percent all SNPs to a minor allele frequency of 2.5 percent in European and Asian populations, and 76 percent of all SNPs to a minor allele frequency of 2.5 million in African populations.
Based on that coverage, Illumina's planned Omni5 should provide coverage of the content in the 1000 Genomes Project upon its release. "I think we can tag 80 percent of those 30 million to 35 million SNPs with about 5 million SNPs in all three populations," Rosenow said of the Omni5 array.
Rosenow said that one myth the firm is trying to dispel is that by looking at rarer human variants, customers will need to look at far larger populations to make any meaningful associations. Instead, Illumina believes that these new chips might enable customers to run fewer samples to power their studies.
"Any GWAS customer, anybody who has run a successful or unsuccessful GWAS" would be interested in the Omni2.5 or Omni5, Rosenow said. By applying a "new theory of identifying rare variants with higher penetrance or effect size," customers "can or might be able to explain more of heritability or what is defined as missing heritability."
Since these variants might have a higher effect size, Rosenow said, customers do not need to run as many samples. Using the chips, he speculated that customers could have a high-powered study with as few as 1,000 samples.
"A lot of customers think they need to run 10,000 samples" using the new chips, Rosenow said. "This is not the case."