Fresh from publishing the results of its first phase this month, members of the MicroArray Quality Control consortium will meet this week in Jefferson, Ark., to map out the objectives for the second phase of the MAQC project.
The MAQC is now soliciting a greater presence from pharma in the project, according to Weida Tong, director of the FDA's Center for Toxicoinformatics at the National Toxicological Research Center and an MAQC organizer. While the organization’s current roster of participating organizations includes a Who’s Who of array manufacturers, only two drug developers are on the list — Novartis and Biogen Idec.
The first phase of the US Food and Drug Administration-sponsored project, published in a special issue of Nature Biotechnology two weeks ago, sought to prove that microarray results are both repeatable and reproducible across platforms.
To show this, the group ran the same RNA samples on array platforms provided by Applied Biosystems, Affymetrix, Agilent, GE Healthcare, Eppendorf, Illumina, Telechem, and an array using oligos made by Operon and printed by the National Cancer Institute (see BAN 9/12/2006).
But rather than focusing on microarrays as a tool with an emphasis on input from platform vendors, the second phase of the MAQC project will look to determine whether or not molecular signatures are reproducible and repeatable — an effort that organizers believe will draw greater interest from the pharmaceutical industry.
“The goal of phase II is to get a better understanding of molecular signatures,” Tong said.
“We want to know whether a molecular signature is really reproducible,” he told BioArray News this week. Tong referenced a number of recent papers that “found that molecular signatures are not reproducible” as a touchstone for the MAQC’s work.
“Some have argued that if you change the signature’s size, maybe just drop a few samples, the signature will be totally different” from a previous molecular signature, Tong said. “But the most distracting thing, it has been argued, [is] that molecular signatures are not reproducible and are incapable of being predictive — the main point is that microarrays are not ready [for] prime time to deliver molecular signatures. We want to find out whether or not we can generate a reproducible molecular signature.”
Such an undertaking not only has ramifications for chip sellers, Tong argued, but also for pharmaceutical companies, many of which are preparing diagnostics based on molecular signatures derived from microarrays.
“We see a lot of the voluntary genomics data submissions from pharmaceutical companies where they develop a molecular signature and try to predict treatment outcomes” based on that signature, Tong said. “These types of molecular signatures are quite frequently used in the drug-development stage. So there are issues with FDA on how to evaluate signatures developed by pharmaceutical companies,” he added.
“We need to have a strong presence from pharmaceutical companies as we begin to decide how this project will work,” Tong said.
Still, that doesn’t mean that array companies will be completely sidelined as the MAQC project takes on new objectives.
“Phase II won’t involve the generation of new data, so in that sense it will be smaller. It will be looking at data sets that already exist in the diagnostics space and then looking at algorithms to see how these datasets be treated,” said Shawn Baker, Illumina’s gene expression scientific product manager.
“We definitely intend to be involved. We’ve been in contact with [MAQC] and given them some data. We hope to be involved with it from the beginning just as we were in the first phase,” Baker told BioArray News last week.
Meeting in Jefferson
Members of the MAQC project are scheduled to meet at the NCTR’s headquarters in Jefferson, Ark., on Sept. 21 for their first face-to-face meeting since the project gathered to discuss the results of phase I in Palo Alto, Calif., in December 2005.
According to Tong, this phase will look to use existing data rather than to generate new data.
“We are asking the community to nominate data and then we will have a data committee to set up the criteria and agree with the consortium to select which data we will use for this second phase of the MAQC project,” he said.
“The first step is to select the right data set, the next step is to determine what method we will use, what data we will use, [and] what questions we will answer. And then we will go from there,” Tong explained.
“We need to have a strong presence from pharmaceutical companies as we begin to decide how this project will work.”
According to Federico Goodsaid, a staff scientist in genomics at the FDA’s Center for Drug Evaluation and Research, the MAQC will now look to evaluate models for selecting biomarkers as well as the reproducibility of signatures.
“There are many different models that are currently in use to select the sets
of genes that will be used as biomarkers,” Goodsaid told BioArray News last week. “We would like to investigate these models and to understand the differences between them and what is required to come up with reproducible signatures to work with,” he said.
Goodsaid said that the MAQC will also seek to answer questions like, “How should a pharmaceutical company identify a set of genes that can be used as prognostic tools? Is there a process more accurate than that from a haphazard choice of models?”
According to Goodsaid, there is “considerable enthusiasm for this particular phase, not only from the platform providers but from a whole new constituency in the MAQC — the pharmaceutical industry.”
“The interest from the standpoint of the platform providers is in terms of applications [that] these platforms will have in the future, all of which depend on the reproducibility of results for these platforms,” he explained.
As for a timeline for when the results of MAQC phase II could be available, Tong noted that the group isn’t even sure what it will be studying yet. However, he said that MAQC has been in discussions with Nature Biotechnology to publish future results as it had the phase I results.
“Hopefully we can publish another paper in Nature Biotechnology,” Tong said. “We have no commitment yet and the project has not started yet, but it is our intention to publish,” he added.