Affymetrix recently announced that t would make eight new GeneChip arrays available in 2005 through its Consortia Program, which the firm launched a year ago at the International Plant and Animal Genome Conference. The arrays will cover canine, Rhesus macaque, legume, Brassica, tomato, citrus, poplar, and sugar cane. Last year, Affymetrix launched grape, soybean, wheat, rice, maize, cow, chicken, and pig arrays as part of the program’s initial offerings. The program is somewhat similar to one launched in January by Agilent, though one of the major differences between the programs is that the Agilent initiative is strictly a custom, not a catalog, array operation. Lianne McLean, director of gene expression marketing for Affymetrix, explained in an interview with BioArray News this week the process for participating in Affymetrix’s Consortia Program.
How does the process work for the GeneChip Consortia Program?
We’ve worked with community groups in the past. We’ve done this for designs that are in our catalog already. For example, Arabidopsis was approached this way, pseudomonas and drosophila. But what we found with the custom program, was that when the large community groups got together and tried to pool resources and tried to do all the sequence selection themselves, it often took a long time for that work to be done. For the larger groups what we found was that if we put together a program that helped them with sequence selection, helped organize the communities by having conference calls, going through a list of standard design options, letting them make the decision about what goes on the array, then underwriting the design fee, these designs could be done a lot quicker. What we’ve done is basically established a number of slots per year, talked to the various community groups, [and] got a sense of how many arrays they would order. If there’s a minimum threshold, and there’s a strong community and enough sequence information to do the design, then we will fit it into those number of slots. If there’s more designs, we’ll put them in the following year.
Who decides what designs will be considered for the program, and who decides the probe design of these chips?
What we’ve done is gone through our sales team and our support team and inquired with the different community groups about where they are. A lot of the groups are interested in doing the design, but they’re just not ready. For example, they’re still doing sequencing. There are groups we’ve been in contact with, but we know they’re not ready, and they won’t be ready until 2006 or 2007. We have an ongoing list of folks, and we touch base with them in the later half of the year. We looked to see where they were with their sequencing, when they’d be ready, how much information they have, how many arrays their community group could commit to, and then we made the decision from there. What actually goes on the design is absolutely dictated by the community.
Would these be considered custom chips, or do they wind up being catalog arrays?
They end up being in the catalog, because we actually do the sequence selection. We put all of our standard controls on these chips, and we do all the actual hard design work. The customer directs us to the sequence they want on the chip. It’s been an interesting process, because we require that the sequence that goes on our catalog chip be in the public databases. So, what’s happened in our discussions with these community groups is [we] encourage groups worldwide to add their sequences to the public databases. If you went back and looked at, say, the barley chip, when we first started discussion with the barley community — and that was custom and was before the program — there were tens of thousands of sequences in the public databases. By the time all the discussions had started and all the worldwide participants had gotten together, everybody committed to putting their sequences into the public databases so they would be represented on the chip … now there’s 450,000-plus ESTs in the database. It’s been really effective in helping drive information into the public databases and making that data available to everybody. At the end, Affymetrix creates a tool that’s available for everybody worldwide.
So, these arrays would be available to everybody, not just consortia program participants?
That’s right. The community groups can be very organized within their communities — for example, the soy group has its own functional genomics committee, and that group assumed the leadership role of communicating with the community at large and helping us on the design process. Other groups don’t have that level of organization. It’s more ad hoc. The people who are interested in the array participate, but we try to reach out to everybody and make sure they’re aware of what we’re doing so they can participate as much as they want. Some people want to direct what goes on the array and be involved in the prioritization discussion, [while] some people just want to buy the arrays when its done. Some of the other groups that are working on some of the pathogens associated with model organisms have been able to weigh in. For example, on the chicken array there are several chicken viral sequences that are included. It’s been really effective in making sure that this is going to appeal broadly.
As part of the program, Affymetrix underwrites the design fee. Can you tell us how much that fee usually is?
Depends on how many arrays the customer purchases, because the design fee is on a sliding scale. It can be anywhere from zero up to $200,000, depending on how many arrays the customer buys, and depending on what format the custom array is in.
Are there any other cost-related incentives for consortia to participate?
I think eliminating the design fee gets rid of the biggest hurdle. Affymetrix donates in-kind work, because we’re offering our sequence selection time as well as our standard probe selection time. So, these groups don’t have to do any of that work. [Also,] the way our contracts are set up is volume-based by institution. So, if an institution is buying a large number of arrays across the institution, then the consortium members of that institution would get a price benefit as well.
Do you have to deal with any of the contracting and legal issues related to sharing intellectual property?
If a customer is doing a custom design, they own that design and they can determine who can have access or if everybody has access. With the consortia program, this is something we set up where we initiate discussions with the community groups, we design arrays for those community groups, and we make those available to everybody. If a customer does a custom design and wants it made available to everybody, we’re happy to do that. They just have to give us the rights to distribute their design to everybody … we’ve been doing that for a long time.
Are there any plans to expand this program?
Well, we’ve been really pleased with the success of this, and we’ve gotten good feedback from some of the granting agencies that this is the approach they would like to see with industry working with academics. So, we’ve been pleased about that. I see us continuing this program as long as it’s successful, and it has been successful so far.