After years of outsourcing its array-based genotyping needs to a variety of players in the market including Affymetrix and Illumina, Merck Research Laboratories, the R&D arm of the pharmaceutical giant, has made the decision to bring the technology in house.
Alan Sachs, vice president of molecular profiling at Merck, told BioArray News last week that the company has decided to invest in its own genotyping capabilities. Though he declined to name the company that will be providing MRL with the equipment or to outline the kind of facility the firm will build, Sachs mentioned both Affymetrix and Illumina as possible vendors. He also said Merck will continue to outsource some of its genotyping needs.
Merck and Affy subsidiary Perlegen have had at least one R&D alliance in the past while Illumina has not disclosed whether it has ever partnered with the drug maker (see BAN 10/5/2005).
Merck's decision to open up a genotyping lab within MRL is a change from its earlier strategy of outsourcing all of its genotyping needs to Perlegen, Affy, ParAllele, and Illumina. It also comes at a time when other pharmas and biotechs, such as Roche and DeCode Genetics, are investing more in array-based genotyping equipment.
Sachs said that the company decided to build an internal genotyping facility because the technology has matured and Merck believes that making a greater investment in genotyping will ultimately aid its drug-development programs.
"We were arguably conservative in our willingness to bring in any one of these platforms until the time was right, and we didn't feel the time was right until recently."
"The need to differentiate products, and the anticipated need to establish efficacy in the correct population is an ever increasing pressure," Sachs said.
"Any tool that enables our company to win at differentiation and choose the right population, from the patient's perspective and the payor's perspective, is a tool that we need to advance as quickly as possible," he added.
According to Sachs, Merck tracked and evaluated the technology until it decided it was at "the right cost and the right scale." Before it decided to bring the technology in-house, Merck "did not feel that the time was right to bring in the platform, and that the technology and the cost and the version that [it] would be acquiring wasn't mature enough," Sachs said.
"We run a very large-scale gene-expression facility, so probably more than any other pharma, we understand what it takes to do this at a large scale and what happens when you commit to a capital investment like bringing in genotyping," he explained.
"And so we were arguably conservative in our willingness to bring in any one of these platforms until the time was right, and we didn't feel the time was right until recently," he said.
Sachs stressed that the company will continue to outsource some of its genotyping needs.
"Our model will be infrastructure sufficient to having throughput of some number and we will be looking for partners externally to provide additional capacity that we may need," he said.
A Domino Effect?
Merck's decision comes at a time when some pharmaceutical and biotechnology companies are reevaluating the way they implement array-based genotyping into the discovery process.
According to Mitch Martin, who heads genomic research at Roche's Nutley, NJ-based R&D center, recently "there has been a greater willingness to consider whole-genome genotyping in applications."
Martin told BioArray News in May that the publication of the Human HapMap project coupled with the availability of a number of high-throughput genotyping technologies that "are getting better all the time and cheaper all the time" had prompted Roche to adopt the technology more widely in its discovery process.
"Previously, Roche used [genotyping] mainly in research applications in doing screening of the genome and of candidate genes to look for associations to disease susceptibility," Martin said.
"We now begin with human clinical information instead of just knockout information in a mouse model. The high-throughput genotyping technologies like Illumina, Affy, and Perlegen can help with the stage of drug target identification and assessment," he added.
Like Merck, Roche will couple capital investments in genotyping equipment with a continued policy of outsourcing some projects to Affy, Illumina, and Perlegen.
"The need to differentiate products, and the anticipated need to establish efficacy in the correct population is an ever-increasing pressure."
"Typically at Roche, the tradition has been to do things in house, with respect to a lot of technologies," Martin said. "[But] we are mixing."
One company that has disclosed its decision to bring genotyping in house has been DeCode Genetics. The deal, concluded with Illumina in May, calls for Illumina to install a customized suite of its SNP-genotyping BeadStations at DeCode's labs in Iceland (see BAN 5/16/2006).
In a statement at the time, DeCode CEO Kari Stefansson said the alliance aims to accelerate the company's target-discovery work. "Our recent discoveries have demonstrated the power of applying high-density SNP genotyping to our population resources," Stefansson said.
DeCode could not be reached for comment. Illumina CEO Jay Flatley told BioArray News last week that the company expects DeCode's installation to be operational by the end of this month.
During a conference call discussing Illumina's first-quarter results in April, Flatley said that pharma has changed its attitude towards genotyping (see BAN 4/25/2006).
"As a group they've moved from the consideration of whether to do whole-genome genotyping to how to do whole-genome genotyping," Flatley said at the time. "There [now] are increased considerations over which clinical trials to use it in and whether to outsource it or to bring it in house."
Last week, Flatley said that since April his views on pharma "have been reinforced."
Affymetrix could not be reached for comment and Perlegen declined to comment.
While pharmas and biotechs are growing more comfortable with array-based genotyping, both Merck's Sachs and Roche's Martin said that while the technology looks promising it has yet to pay off.
"This is the very, very first step in the pipeline and the pipeline is a 10- to 12-year endeavor. It's really going to take some time before we can say with any certainty whether or not it's paid off," Martin said.
Martin did say that genotyping work was not only helping Roche find new biomarkers, but backing up some of its earlier discoveries.
"The pattern that is coming out of all this work â€" at least in our heads â€" is lighting up not only novel genes that we would have never considered as potentially interesting drug targets, but ones that are already known drug targets. Those are coming up â€" and that suggests that yes, we are on the right track and this is working," he said.
"But I think overall, it's still early days," Martin added.
Similarly, Sachs said that, at the moment, Merck is using genotyping "very much in the discovery mode." He said that current work will be followed up with custom assays once Merck starts testing hypotheses.
Affy and Illumina can count as early adopters Merck, Roche, DeCode, GlaxoSmithKline, and Pfizer, which bought a 12 percent stake in Perlegen last year. But there are others on the fence.
Mary Klem, a spokesperson for Amgen, told BioArray News last month that the company is interested in using tools like those sold by Affy and Illumina, but that it has yet to significantly invest in genotyping.
"While Amgen scientists are following the development of SNPs with interest, the application at Amgen today is relatively limited," Klem said.
â€" Justin Petrone ([email protected])