When microarray technology came onto the market in the late-1990s, the high price tag pushed it beyond the reach of most academic labs, so researchers were forced to use their initiative. At the time, Stanford University's MGuide on how to build your own arrayer from scratch proved invaluable. Today, there are many more options.

Although it is possible to build an arrayer for about US$50,000, a basic instrument can now be bought for about the same price from a number of suppliers — such as BioRobotics of Cambridge, UK, Genetix of New Milton, UK, Cartesian Technologies of Irvine, California, and GeneMachines of San Carlos, California — although prices for arrayers vary widely depending on their speed and capacity.

At the same time, the GeneChips made by Affymetrix are now more affordable; and ink-jet systems are starting to trickle onto the market, offering greater speed and more uniform spot morphology over contact printing, but these still come with a fairly hefty price tag.

The demand for the technology is so great at some of the major research institutions that they have established core facilities to produce inexpensive microarrays and so make the technology more broadly available. Some facilities with spare production capacity are also selling arrays at cost to investigators from outside institutions. Stanford University's Stanford Functional Genomics Facility, for example, offers human microarrays containing 49,000 cDNAs, of which 15,000 or more are unique human genes. The KIChip core facility at the Karolinska Institute in Stockholm offers a range of services to external researchers on a fee basis, including the production of custom-spotted microarrays.

When Vivian Cheung, of the departments of neurology and pediatric oncology at the University of Pennsylvania, Philadelphia, first thought about using microarrays in late 1996, there were no commercial arrayers and GeneChips were out of her price range. Her only option was to build a DNA arrayer in-house. Cheung's SPOT DNA arrayer has churned out arrays for her lab for several years, where one of the main aspects of research is narrowing down the location of genes responsible for genetic diseases. She still makes and reads her own arrays, but has switched to a commercial instrument from Affymetrix. “The price is coming down to a point where it's worth our while to buy the instruments and have someone else take care of them,” she says.

On the other hand, Michael Miles, of the department of pharmacology and toxicology at the Virginia Commonwealth University in Richmond, admits to being “sort of biased towards the commercially available arrays”, which he has been fortunate enough to be able to afford. Miles is studying the molecular plasticity of drug abuse and says DNA array studies provide a genomic-level, non-biased approach. He buys commercially produced chips but processes them on his lab's Affymetrix scanner, still a fairly expensive item at just under $200,000.

Data analysis at the Ontario Cancer Institute's Microarray Centre. Credit: ONTARIO CANCER INSTITUTE

Whether it makes sense to buy off-the-shelf or make your own arrays also depends on how many you need, and whether commercial arrays contain the genes you are interested in. But doing it yourself is not always easy. Jan Vijg, of the department of physiology at the University of Texas Health Science Center, San Antonio, says it took endless telephone calls, numerous lab visits and more than a year to develop a workable system. His research interests centre on the molecular basis of ageing and cancer. In 1999 he looked into buying commercial arrays but realized that to do large-scale experiments of 100–200 arrays at a time would mean making his own arrays in-house. “Everything we did is really based in one way or another on information in the public domain,” says Vijg. He bought a BioRobotics arrayer and an Axon Instruments scanner, and adapted the Stanford protocols, initially printing 2,000 genes per slide in duplicate. He has since been asked to turn his facility into an institutional core and has bought a second arrayer, this time from GeneMachines. This comes with a price tag of $120,000–130,000 but has better throughput and capacity. “I think eventually we'll be able to make 20,000-gene slides in duplicate available for less than $100,” says Vijg.

When Jim Woodgett began to dabble in microarray technology three years ago, he never expected to end up running a core facility that supplies high-density microarrays and technical support to academic researchers across the globe. The Microarray Centre at the Ontario Cancer Institute in Toronto, which he directs, was established through a partnership between the institute, the government and industry to ensure that Canadian scientists had access to affordable high-quality microarrays.

Woodgett contracted with Toronto-based Engineering Services, now Virtek Vision International, to design a contact arrayer that uses a split-pin configuration. The company now sells a third-generation version of the original. Woodgett's centre runs four machines in parallel, printing from 48 genes at a time. The 30 staff generate the probes, produce the arrays and carry out quality control, and include technicians, researchers and bioinformaticians. Last year they produced 16,000 off-the-shelf arrays, including human and mouse arrays — most of which were high density, and 40% of which went to academic labs outside Ontario, many to the United States.