Affymetrix and the UK Biobank last week announced an agreement that will see the array vendor genotype UKBB's collection of 500,000 samples.
UKBB collected the samples from participants between 2006 and 2010. Along with providing blood samples, the participants, aged between 40 and 69 years, also submitted urine and saliva samples, as well as information about themselves. They also agreed to have their health assessed by the project in the future.
UKBB said it is working with Affy to better understand the relationship between lifestyle and genes in causing major diseases that affect the UK population such as heart disease, dementia, and cancer. In total, UKBB has raised £21 million ($32 million) to support the genotyping, which Affy will perform. Affy said that it expects to begin delivering data for the study by the end of the year, and that it expects to recognize the majority of the revenue connected with the project in 2014.
According to Mindy Lee-Olsen, vice president of marketing services at the Santa Clara, Calif.-based company, the partners will use a custom Axiom Biobank Genotyping Array to genotype the 500,000 samples.
Affy launched a catalog Biobank array containing about 650,000 variants last year, and has also allowed its customers to design their own chips based on the catalog product (BAN 10/16/2012).
Lee-Olsen told BioArray News that data from UKBB's study will be made available to other researchers. She declined to provide specific array design details.
When Affy launched a catalog version of its Biobank array last year, it said it was hoping to tap into perceived demand for "cost-efficient" arrays from biobanks, genome centers, and core laboratories that aimed to conduct genome-wide association studies using their holdings.
Enticed by the same opportunity, Illumina also launched a new family of HumanCore arrays around the same time (BAN 10/16/2012).
Apparently, the two vendors were right. Illumina CEO Jay Flatley told investors during the San Diego firm's fourth-quarter earnings call that Illumina received orders for enough HumanCore arrays to process 325,000 samples between the introduction of the arrays in mid-October and the end of the year (BAN 1/29/2013).
Mike Nemzek, vice president of strategic marketing at Affy, told BioArray News this week that, since Affy shipped its first Biobank arrays in Q4 of 2012, the "initial interest and level of purchases for the product are the strongest we have see for any Axiom Genotyping Array to date."
Mark McCarthy, a professor at the University of Oxford who helped designed Illumina's HumanCore array, said that he is "not entirely surprised" by the "substantial" uptake of the vendors' biobank arrays.
"There were many large and valuable biobanks that had missed the earlier GWAS train, and for whom the cost profile of these arrays, and the addition of exome content, has provided a valuable opportunity to get on board," McCarthy told BioArray News this week. "The availability of ever better reference panels … and the still-high cost of whole-genome sequencing, means that it is increasingly important not to miss out on having a good GWAS scaffold," he added.
Cost is certainly a factor that favors biobank arrays. But, according to McCarthy, the adoption of these arrays is part of another change that favors biorepositories with larger, more diverse sample collections.
"We are seeing a shift away from disease-specific case-control resources, though they remain important, to large cohorts with health record linkages, such as UK Biobank or Kaiser Permanente," said McCarthy. Affy inked an agreement to provide Kaiser with arrays to genotype 100,000 samples as part of a longitudinal health study in 2009 (BAN 10/20/2009).
Apart from scale, McCarthy said that there are several ways in which genotyping such cohorts could provide researchers with "powerful translational opportunities." One is the ability to explore the consequences of a given variant, and its effects on the function or expression of its cognate transcripts, across a wide range of phenotypes. That, in turn, enables scientists to better "understand pleiotropy, [when one gene influences multiple traits], and better prioritize potential therapeutic targets for those unlikely to be contaminated by an adverse risk profile," McCarthy said.
Still, McCarthy cautioned that, in comparison to the UKBB project, smaller biobanks might not be able to benefit from the opportunities to customize their arrays to the extent that UKBB did, given more limited resources. Queries to several other biobanks, including Generation Scotland and the Estonian Genome Project, were not returned. Generation Scotland announced in 2011 that it was working with Arrayjet, a UK-based microarray instrumentation and services provider, on biomarker discovery.
"The specific tailoring of the Affymetrix array that was possible for UKBB was in part enabled by the massive scale of the project, and it remains to be seen whether equivalent redesigns are possible for smaller projects," he said.
At the same time, he noted that both vendors allow custom content to be added to existing catalog arrays, so he expects "this trend of customization" to continue.
"It will be particularly interesting to see whether the vendors are responsive to the need for optimization of content to cater to a wide range of major ethnic groups, to optimize imputation properties, and to capture rarer more ethnic-specific variants," said McCarthy. "There will surely be increasing demand from this direction, and it may not be well served by bolting a modest amount of custom content on an array otherwise designed for another major ethnic group."