Skip to main content
Premium Trial:

Request an Annual Quote

ABRF Survey Finds Affy Chips, Gene Expression Still Dominate Array Field

Premium
SALT LAKE CITY — Despite the influx of new applications for microarrays over the last several years, core labs are still running more arrays for gene expression studies than for any other application, according to preliminary results of a recent survey by the Association for Biomolecular Research Facilities’ Microarray Research Group.
 
The 2008 MARG survey also found that Affymetrix GeneChips are the most commonly used array platform, followed by Agilent arrays, in-house spotted arrays, and Illumina’s BeadChip platform.
 
Furthermore, the study found that in spite of the recent rise of next-generation sequencing technologies — which some consider to be a promising alternative to microarrays for many research applications — only 9 percent of respondents said they believe the sequencers will replace microarray technology in the next five years.
 
MARG member Christina Harrington, director of the Affymetrix Microarray Core at Oregon Health & Science University, presented the preliminary results of the study on Sunday at the annual ABRF conference here. She said that the group expects to publish the complete results on its website next week.
 
The survey was conducted last November and December and had 149 respondents. Of those, 77 percent were academic or government microarray facilities and the remainder were commercial service facilities or pharmaceutical labs. The majority of respondents — 74 percent — were based in North America.
 
The study was the fifth such survey that MARG has conducted since 1999, with the most recent one completed in 2005. The 2008 survey was the first study to include array platforms other than Affymetrix GeneChips and custom spotted arrays.
 
Susan Hester of the Environmental Protection Agency’s Cellular Toxicology Branch, who serves as MARG chair, told BioArray News that the 2005 survey gave labs the option to select commercial array platforms other than Affy, but no respondents said they were using those systems at the time — a reflection of how rapidly the field has changed over the past three years, she noted.
 
Even though firms like Illumina were selling expression arrays in 2005, they were still relatively new to the market, and Affy had the first-mover advantage at that time, Hester said.
 
Harrington said in her talk that new applications and new technologies that have emerged since the 2005 survey have led to “much discussion” in the microarray community about where the field is going, and that one aim of this year’s study was to provide a snapshot of the current state of the art, as well as the future of the technology. 
 
The Affymetrix platform is the most widely used, according to the survey, with around 60 percent of respondents noting that they run Affy GeneChips. Agilent arrays and in-house spotted arrays were the next most popular, with about 50 percent of labs saying they used those platforms. Around 30 percent of respondents said they are using Illumina BeadChips, while fewer than 10 percent said they are using arrays from Applied Biosystems, NimbleGen, or CombiMatrix.
 
In the 2005 survey, 97 labs, or 55 percent of the 177 respondents, said they were using spotted-array technology, while only 45 percent said they were using the Affy platform.
 
Both Harrington and Hester told BioArray News that they were surprised that so many labs are still running custom spotted arrays in light of the rise in commercial catalog chips over the past three years. Harrington said that many labs prefer to use in-house arrays for targeted assays of a “unique set of genes,” or for custom studies involving a mixture of species on a single chip. Many labs appear to be spotting their own arrays for microRNA studies as well, she said.
 
In terms of throughput, 29 percent of respondents in the current survey said they process fewer than 200 arrays per year, 36 percent said they process between 200 and 1,000 per year, 18 percent said they process between 1,000 and 2,000 arrays per year, and 18 percent process more than 2,000 arrays per year. Harrington stressed that these estimates are rough, since some labs may count several arrays on a single slide as one array.
 
Most respondents expect their throughput to increase in 2008: 52 percent said they expect to run more arrays in 2008 than 2007; 38 percent said they expect their throughput to remain at about the same level; and 11 percent said they expect a decrease in throughput this year.
 
It appears that labs in general are doing more work with less staff, however. The average number of staff per lab has dropped to 3.2 in 2008 from 3.8 in 2005 and 4.7 in 2003, and the majority of core labs now have fewer than three staffers, the study found. Furthermore, only 32 percent of respondents said they expect to add staff in 2008, compared to 52 percent who planned to add staff in the 2005 survey.
 
Gene Expression Still on Top
 
The study found whole-genome expression profiling is still the most widely used application for microarrays. Around 98 percent of Affy customers said they are running gene-expression GeneChips, while 86 percent of Illumina customers run gene expression studies on the BeadChip platform.
 
Around 29 percent of Affy users and 27 percent of Illumina users run whole-genome genotyping studies with more than 1 million markers, while 44 percent of Affy users and 60 percent of Illumina users run whole-genome genotyping studies with fewer than 1 million markers. 
 

New applications and new technologies that have emerged since the 2005 survey have led to “much discussion” in the microarray community about where the field is going.

GeneChips appear to be the most popular arrays for gene-expression analysis in most labs, according to the survey. In response to the question, “What arrays do your investigators prefer to use for gene expression analysis?” 48 percent of respondents said Affy arrays. Agilent arrays are used preferentially for gene expression in 21 percent of labs, Illumina arrays in 20 percent of labs, and in-house spotted arrays are used in 16 percent of labs. Fewer than 2 percent of respondents said they prefer to use arrays from ABI, NimbleGen, or Exiqon for expression studies.
 
Asked to look forward, respondents cited splice variant arrays as a particular area of interest. Around 68 percent said they are “interested” in splice variant applications, half of whom are already running the chips.
 
Another area of interest is clinical applications, with 50 percent of respondents noting that they are thinking about running arrays on clinical samples. However, only 38 percent of this number have investigated the regulatory guidelines for running such arrays, and only 10 percent said they were familiar with the National Institute for Standards in Technology’s External RNA Control Consortium program addressing requirements for running arrays in the clinical setting
 
NIST is currently testing the 96 synthetic mammalian RNA spike-in controls that it will recommend for gauging the performance of microarray assays. The controls should become available through NIST sometime this year (see BAN 7/17/2007).
 
Grappling With Array Data  
  
Echoing the findings of the 2001, 2003, and 2005 MARG surveys, the 2008 questionnaire found that data analysis remains the primary challenge for microarray labs. Around 65 percent of respondents cited bioinformatics among the main challenges they face, with operations funding being the next most-pressing issue as 40 percent of respondents called it a key challenge.
 
Around 45 percent of respondents said they have one or more full-time employee dedicated to analyzing array data, while 23 percent said they outsource array analysis to a contractor.
 
More than 50 percent of respondents said they use software other than what was provided by the array vendors for processing and analyzing arrays. 
 
The most popular software platform cited for array analysis was Agilent’s GeneSpring, which is used by 60 percent of responding labs. Other popular analysis packages include the open source R statistical package, used by 45 percent of respondents, and Microsoft’s Excel spreadsheet software, used by 30 percent of respondents.
 
For analyzing the biological significance of array results, around 50 percent of responding labs said they use Ingenuity Pathway Analysis; 35 percent said they use the National Institute of Allergy and Infectious Diseases’ Expression Analysis Systematic Explorer (EASE) or DAVID (Database for Annotation, Visualization and Integrated Discovery) tools; and 33 percent of labs said they employ Ariadne Genomics’ Pathway Studio, formerly called Pathway Assist.
 
Harrington said that the study found “no widely used, highly rated database” for managing array data among the labs surveyed. Around 47 percent of respondents said they store their data in some form of database, including systems like the open source BioArray Software Environment (BASE), the Food and Drug Administration’s ArrayTrack, and Rosetta Resolver, but there was no consensus among respondents regarding the best system for storing and accessing microarray data.
 
Around 13 percent of respondents said they “always” submit their experimental data to a MIAME-compliant database such as the Gene Expression Omnibus or ArrayExpress, while 21 percent said they “sometimes” submit data to these resources and 39 percent said they submit only when it is required by a journal publisher.
 
Of those that submit to these databases, 84 percent deposit data in GEO and 31 percent deposit data in ArrayExpress.
 
The Impact of Next-Gen Sequencing
 
Harrington said MARG is particularly interested in finding out what sort of impact high-throughput sequencing technologies are expected to have on microarray facilities, and whether core labs think those systems are likely to displace microarrays down the road.
 
“In essence, the answer was no,” she said.
 
Only 9 percent of respondents said they think next-gen sequencers would “completely” replace microarrays within five years, although 62 percent said they think the technology would “partially” replace arrays. Around 17 percent of respondents said they think next-gen sequencing would have no impact on arrays at all, while 12 percent said they are unsure.
 
Of the responding labs, 28 percent, or 23 facilities, currently have a next-gen sequencer: 10 Illumina Genome Analyzers, nine 454 systems, and seven ABI SOLiDs. Around 46 percent of respondents said they plan to purchase a high-throughput sequencer in the “near future,” however.
 
Hester told BioArray News that MARG is keeping a close eye on the adoption of next-gen sequencers in microarray core labs, and said the group is considering working with other ABRF research groups in the future in order to “better assess that technology and see what’s on the horizon there.”

The Scan

Shape of Them All

According to BBC News, researchers have developed a protein structure database that includes much of the human proteome.

For Flu and More

The Wall Street Journal reports that several vaccine developers are working on mRNA-based vaccines for influenza.

To Boost Women

China's Ministry of Science and Technology aims to boost the number of female researchers through a new policy, reports the South China Morning Post.

Science Papers Describe Approach to Predict Chemotherapeutic Response, Role of Transcriptional Noise

In Science this week: neural network to predict chemotherapeutic response in cancer patients, and more.