Over the next few months, Affymetrix plans to debut multiple population-specific genotyping arrays designed to study diverse human populations including Europeans, Asians, and Latin and African Americans, according to a company official.
Candia Brown, Affy's director of strategic marketing for genotyping, said the company will launch five population-themed arrays before the end of the summer. The chips, four of which were co-designed with the University of California, San Francisco, and Kaiser Permanente, will debut in stages.
Next month the Santa Clara, Calif.-based company plans to release an Axiom genotyping array optimized for studying European populations. It will follow this before the beginning of September with Axiom arrays focused on East Asian, Latino, and African American populations.
Additionally, Affy is preparing to launch a high-density array during the summer that will cover genetic variation found in African populations.
The new arrays will expand the company's menu of population-optimized chips. Earlier this month, Affy introduced its Axiom Genome-Wide CHB 1 Array, which is designed to characterize the genetic basis of disease in Han Chinese populations (BAN 4/12/2011).
And last year it introduced its Axiom CEU and ASI arrays for European and Asian populations, respectively. If the company meets its deadlines, it will soon have eight population-optimized arrays on the market.
"With our first two arrays we focused on where the vast majority of GWASs [were] done — in the European and Asian populations," said Brown. "Since then we have had more content released into [the] public domain, so we have expanded our offering to accommodate more genomes and the rare variants that are available."
According to Brown, Affy's population-optimized chips typically comprise three constituents: common variants, common and rare variants, and very rare variants suspected to be linked to disease.
"We recognize that many of our customers utilize different design strategies to conduct their association studies," she said. Some are "still focused on the common variant/common disease model," while others are more focused on looking at both rare and common variants to study gene-gene interactions or at more "clinically actionable" SNPs and genes.
"We are developing a strategy that allows us to accommodate customers no matter where they are entering into their strategy, whether they are focused on common variants, rare variants within 2-percent [minor-allele frequency], or they really want to focus on the really rare and potentially causal variants down to 1-percent minor-allele frequency," she said.
The recently launched CHB 1 Array, for instance, falls into the first category of chips, while the Axiom CEU and ASI Arrays that were launched last year are in the second.
The four arrays developed with UCSF and Kaiser Permanente, though, will focus on very rare variants and fall into the third group of arrays: very rare and potentially causal variants.
"Together with UCSF, we have developed a suite of population arrays that is more hypothesis-driven and it is more focused on genes or SNPs that have been implicated in disease," Brown said. "Those include rare variants down to 1-percent minor-allele frequency."
As part of their effort in the project, Kaiser and UCSF researchers aim to use custom Axiom arrays to genotype 100,000 volunteers by the middle of this year. The scope of the project has enabled the researchers to develop custom arrays to look across the genomes of distinct population groups, an approach they recommend to others.
Pui-Yan Kwok, a professor at UCSF involved in the Kaiser study, told BioArray News last November that, given the output of the 1000 Genomes Project and the Kaiser-UCSF study, most researchers conducting GWASs will run population-specific studies in the future.
"For association studies, you are trying to match cases and controls so that the only difference between them is their disease," Kwok said at the time. "But if you don't match them with the ethnic background properly, then you will be getting hits the wrong way and you will be overwhelmed by the structural differences between the two populations" (BAN 11/9/2010).
To meet that demand, Affy will next month launch a European array designed in part with the UCSF-Kaiser researchers. Debuting early this summer, meantime, will be an East Asian array that will "address multiple populations and or [mixed] populations of Asian descent."
By the end of the season the company will launch the African American and Latino arrays. All the offerings are designed to survey up to 800,000 SNPs per array, said Brown.
[ pagebreak ]
"These are very hypothesis-driven, focused on variants that are already known or genes that have been implicated and focused on their putative association with disease," Brown said of the four new arrays. "We are making them available as UCSF completes screening across that portion of their cohort. When it is done we [will] look at the data quality and make those arrays available to other researchers."
Brown noted there has been "significant demand" for the new chips ahead of their launch. One reason is their "very rare" variants, while another is UCSF and Kaiser's contribution of "phenomenal" phenotype-genotype information.
"That will help customers to correlate genotype phenotype information, and especially as it relates to their Latino Array and African American Array, as there has never been an array designed to address variation in Latino populations," she said. "This will be very novel for them."
'Highly Complex' Africans
Then there's Affy's plan to debut a high-density chip for studies of African populations. The chip, called the Axiom Yoruban Array, will comprise about 2.3 million SNPs and cover around 90 percent of the Yoruban population and "all variation across the African genome," Brown claimed.
Affy plans to commercialize it this summer.
According to Brown, the genomes of Africans are "highly complex" because they are older and have been through more evolution and environmental pressures.
"You tend to see smaller [linkage disequilibrium] blocks, more complex genomic structures, and there is a different degree of minor allele frequency," she said.
The Yoruba are a West African ethnic group located primarily in Nigeria. Contents of the Axiom Yoruban chip came from 30 adult-and-both-parents trios from Ibadan, Nigeria, who were genotyped between 2005 and 2009 as part of the International Human HapMap Project.
Expanding the Axiom
Affy plans to develop and debut other population-optimized arrays in the future.
"As more content becomes available from 1000 Genomes and other populations, as well as other [principal investigators] that are funding sequencing projects in-house, there will be more content available, and we will use that to develop [the] next-generation of population-optimized arrays [and] also to expand our Axiom database," said Brown.
She added that the company will soon launch the second version of the Axiom database, which will enable researchers to access 10 million variants with which they can design new custom arrays. Affy last updated the database in March, when it contained 5.4 million variants (BAN 3/8/2011).
The second version of the database has "much more content across the major populations," said Brown. "We will be enabling PIs to design their own arrays, and if there are groups of researchers looking to standardize and develop consortium arrays we will commercialize those, too."
It is unclear how Affy's main rival in the GWAS market will compete against the firm's growing menu of population-optimized arrays.
An Illumina spokesperson told BioArray News this week that the San Diego company is "always exploring the value" of developing population-specific arrays. "Our current array products have good coverage of most major populations, including Asia-specific populations," he added. The spokesperson also claimed that the firm's Omni2.5 BeadChip has the "best coverage for the Chinese population on the market today."
Have topics you'd like to see covered in BioArray News? Contact the editor at jpetrone [at] genomeweb [.] com.