AT A GLANCE
• Trevor Hawkins
• February 2002-present — Vice President Development, Discovery Systems, Amersham Biosciences
• Previously — Director, US Department of Energy, Joint Genome Institute.
• 1993 — PhD, biochemistry, Cambridge University (UK).
• 1989 — BS, biochemstry, University of Sussex (UK).
Trevor Hawkins, the executive responsible for setting the strategic direction of Amersham’s genomics business, in the wake of its $20 million acquisition of Motorola’s CodeLink microarray business
last summer, said microarray manufacturer did not intend to compete on density.
After Applied Biosystems last week announced that it would bring out a whole-human-genome, single-microarray product to market by the end of the year, BioArray News sought out Hawkins to
speak about density and a heated-up marketplace.
How do you analyze the need for a single whole human genome microarray?
I do not believe that having the entire genome on one microarray is a huge advantage when you are looking for quantitative and reproducible information. We are looking at increasing density, the same as any of the microarray companies, for the human genome and other complex genomes where the number of genes is 30,000-plus. An increase in density is something that is attractive but not necessarily of the highest priority. We have taken the stance that the No. 1 issue, the biggest problem that customers have, is that you don’t want to be buried in data. They want highly quantitative, reproducible, and sensitive information. We have centered CodeLink around those attributes. So, are we looking at a higher density arrays? Yes, it is one of our developing applications. But it is not a race to get it on one chip. If you have high reproducibility on a chip-to-chip system, that’s not much of an issue. So, it is not that we downplay the importance of a whole-genome array, but highly quantitative data, that is our main goal, rather than trying to play the density game of 40, 50, 60,000 probes and what is the limit.
Did ABI’s announcement surprise you?
This is an area that is growing rapidly, and it certainly fits with ABI’s business model to have an offering in this area. So, I’m not surprised. I also think that the competition is good. Choice is good.
What should be on an array?
The question might best be: What shouldn’t be on a chip? We will move to a plan where every and any genome can be on a chip. There are questions on gene expression that you want to ask on every genome — mammalian, microbial, you name it. We will move to that world in the research field. [Any] genome done today will be on a chip. I strongly believe that in the academic world, there will be a market of researchers wanting to look at their particular genome that they have sequenced. We are just now getting to the point where making custom bioarrays that are highly specific, in perhaps more numbers, is not only technically feasible, but is also going to make business sense. We are looking at methods and approaches to build up the CodeLink capabilities.
Do the recent events change the marketplace?
It’s not a big shift. Certainly, the correspondence between the FDA and Roche has been interesting and highlighted concerns that we have had for a while now, about how microarrays can be used in a clinical and diagnostic setting, given what we know about taking anything through the FDA. We know that reproducibility, and manufacturing ability, and lot-to-lot variance are critical. We knew this going into the business. The technology we use to make CodeLink, a methodology that is traceable, the quality metrics that we inherited from Motorola, with Six Sigma [manufacturing protocols], gives us a product that is showing great promise.
The ability to be able to compare experiments in one lab to another, and experiments on one platform to another, even today is very difficult, and standardization is going to be important. What the FDA is really saying is that in order to get FDA approval one needs to show extremely high reproducibility and high quality of results between different batches of your own product and multiple attempts to do the same experiment. The critical thing is the quality, and we recognized that a year and a half ago. Gene expression people are using microarrays as a triage approach. They are looking at a large number of genes, then verifying their methods using RT-PCR. There is discordance in results from RT-PCR and the products that are out there today. That highlights the fact that data you get needs to be highly quantitative, which is critical in clinical and diagnostic arenas.
What does your business look like, going forward?
From day one, we have believed that there are three revenue streams. First there is catalog — human, mouse, rat, P450. The second major stream is custom chips — whether that be for academic groups, where we make it or put it into a catalog, whatever the scenario. Third are slides that allow people [to] use spotting devices that they might have in their labs. [The] microarray market is a good 30, 40, to 50 percent self-spotting. It’s not that transparent in the US, but in Europe, the majority of the market is self-spotting. The US is pre-arrayed, and catalog, but the world still uses spotting devices. We make one and many other companies make others. It’s a big market. Over time self-spotting will become less and less financially feasible, and [it will be] more attractive to make a call to Amersham to have a chip made to the same quality metrics. We will move there. We sell a very large number of CodeLink activated slides.
You are a veteran of the glory days of the rush to sequence the human genome. Any parallels here?
In the early days of sequencing, as the program ramped up, the feeling was that more data was better — produce the data and get it out there. As the program developed, it became obvious that what was created was a database that had real question marks on quality. There was a shift as the genome got to be finished, where quality became the main mantra. We are seeing the same thing now. We are getting a sense that we need higher quality systems, we need to see rare transcripts and measure those using methods that are highly reproducible. The ways the worlds developed are very similar.