COPENHAGEN, Denmark — While next-generation sequencing has become the technology of choice for genomics research, a number of experts who attended Copenhagenomics, held here last week, say that they still use and will continue to use microarrays in their research.
Copenhagenomics is a nonprofit conference focused explicitly on next-generation sequencing technologies. Talks at the conference ranged from the discussion of FarGen, a project that aims to sequence the entire population of the Faroe Islands, to using next-gen sequencing to uncover the impact of circumcision on the penis microbiome.
But many scientists also presented array-related data and projects, and some researchers who mainly use sequencing said they might use the older technology in the future.
"I use arrays all the time," said Vanessa Hayes, a professor of genomic medicine at the J. Craig Venter Institute in San Diego. While Hayes' presentation at Copenhagenomics focused on the use of next-generation sequencing to better understand the genetic relationships between indigenous populations in southern Africa, she told BioArray News that much of the data she presented was array data.
Hayes uses arrays to identify individuals for sequencing, she said. She has used the million-marker HumanOmni1, and recently moved to the 2.5 million-marker HumanOmni2.5 BeadChip. As Hayes is involved in mapping the relationships between isolated groups, she said she needs to be sure that an individual who claims to be from a certain group is actually genetically representative of the population in question.
"It is still very expensive to sequence, and I need to be sure that that person is who I think they are," Hayes said. "In different cultures, there are very few records, and they don't know the historical relationships between each other," she said. "I use the array to figure out who I am going to sequence."
According to Hayes, next-generation sequencing is necessary because most commercial genotyping arrays are "pretty biased" toward European populations, but, she said, using an array is "sufficient" to "put individuals into perspective, how they are related, who are they, so when someone is sequenced, we actually know who they are on the genetic level, because they are so diverse."
While Hayes is using arrays as a selection tool, Umeå University's Douglas Scofield envisions using arrays to study expression patterns in larger populations of individuals. His research is focused on using next-generation sequencing to assemble the genome of the Norway spruce, which he described as the "most complex genome assembly project to date."
According to Scofield, Umeå Plant Sciences Center is coordinating a multinational effort to use RNAseq to assemble sequences from diploid tissue, haploid megagametophytes, multiple transcriptomes, and more than 500 haploid fosmid pools while developing techniques for merging information from all of these sources into a well-annotated genome "useful for tree improvement, functional studies, and basic plant biology."
Still, Scofield told BioArray News the center might use arrays at a later stage in the project. "When you start looking at expression patterns, you might start running into technological limitations where array technology is so good that it makes sense to use that," he said.
"Trying to do RNAseq on a thousand individuals, you are still going to have to deal with assembling a thousand transcriptomes," Scofield noted. "An assembly requires huge computational resources," he said. "Working with a thousand individuals with chip-based technology is not nearly the issue that trying to assemble a thousand transcriptomes [is]."
Still 'Cutting Edge'?
Shawn Baker has witnessed the adoption of next-generation sequencing in genomics research. He said he was Illumina's 15th employee, and held a number of positions at the San Diego company, including marketing manager of expression and epigenetics. Now, Baker is chief science officer at BlueSEQ, an Aarhus, Denmark-based "sequencing exchange" that matches researchers' sequencing service needs with labs that have extra capacity on their NGS instruments.
Baker told BioArray News that researchers continue to use arrays for a number of reasons, including ease of use, installed instruments, and legacy data. He also said that the market continues to demand new chips, citing the success of Illumina's recently launched exon arrays, but that firms that sell arrays are investing more in developing other technologies.
"Arrays are still cheaper, for the moment; people are comfortable with them, the analysis has been worked on for a long time, but even on … the cutting edge of things, you are still seeing arrays being adopted," Baker said. "Illumina said they were surprised that their exon arrays took off as well as they did," he noted. "They made that product, weren't sure what it was going to do, and all of a sudden they got caught off guard because so many people wanted it."
Illumina introduced its exome family of arrays last year. CEO Jay Flatley said during the firm's first quarter earnings call in April that since the launch, Illumina has shipped enough arrays to survey 1.3 million samples. Christian Henry, general manager of Illumina's genomic solutions business, said last month that the market's reaction to the new chips exceeded the company's expectations (BAN 5/15/2012).
Still, Baker said that it is "definitely true" that researchers are switching to sequencing. "We are seeing slowly or not so slowly people move away from array platforms to sequencing," he said. "It is possible that that phenomenon will plateau," he added. "It is not that arrays will go away, they will still always be around for a long time."
A real issue for the future of arrays lies in research and development. "Unless something happens on the protein [array] side, it is unlikely that you will see massive innovation on the arrays," said Baker. "It's not even that they are not a good platform, it's just that there is only so much focus to go around and the traditional array companies are not doing that anymore," he said.
"Illumina still makes arrays, but the vast majority of innovation is going on in sequencing," he noted. "Affymetrix still makes arrays, but they are pushing into that lower multiplex area."
Anne-Mette Hein, a senior scientist at Aarhus, Denmark-based CLC Bio, confirmed that most of the bioinformatics software firm's innovation is focused on sequencing data analysis. "We have some array-related tools and I would say that the downstream tools we have available are applicable to both arrays and NGS," Hein told BioArray News.
According to Hein, the company continues to serve clients who use array technology.
"I think a lot of people have [valuable] experience with that technology; they know what to do with that type of data, and for sure they will keep on using microarrays," said Hein. "Also they have valuable data that they have processed on these microarrays and they can't go back and reprocess those all, so there will be at least for a while the need to combine the two technologies."
'An Honest Hope'
During the conference, there were several recurring themes that might have caused array users to experience déjà vu. One was the issue of cross-platform comparisons, which, in the array space, eventually led to the initiation of the US Food and Drug Administration-hosted Microarray Qualty Control consortium and the first two phases of the group's project, both of which were detailed in several publications (BAN 8/3/2010).
The third phase of the project, called Sequencing Quality Control, aims at assessing the technical performance of next-generation sequencing platforms by "generating benchmark datasets with reference samples," and "evaluating advantages and limitations of various bioinformatics strategies in RNA and DNA analyses," according to the project's website.
But BlueSEQ's Baker warned that it would be more difficult to carry out such projects than it has been for similar, array-focused projects.
"Sequencing platforms are changing so much more quickly than the array platforms that it's really difficult to do these kinds of projects," Baker said. "Groups don't move quickly, you are dealing with a bunch of institutions and companies coming together on borrowed time, and instruments last a year to a year and a half before they are replaced by an upgrade or a new technology," he said.
Baker characterized the sequencing market as "extremely competitive" and said that it has "more active players" than the array market did. "I have a list of between 30 and 40 companies that are trying to make the next, next-generation technology," he said, while he noted with arrays there were only a handful of dominant players, namely Affy, Illumina, and Agilent Technologies.
Another factor that may be familiar to array users is the belief that the new technology will answer most scientists' questions.
"There is a lot of NGS hype now, just as there was array hype 10 years ago," said CLC Bio's Hein, adding that there would be hype about whatever technology is popular in five years. "I think scientists are optimists, they tend to believe in what is promised by the new technology," she said. "With sequencing there is an honest hope that this technology will answer our questions."
Hein acknowledged that sequencing data is "more precise" than array data, but said that a host of other issues make data analysis complex. "We thought that counting reads would be easier than counting signals from an array, but there are others statistical issues that arise," she said. "So I think that sequencing does supply more information, but it is not the end, and there are new technologies that could be better, and more promising."