PALO ALTO, Calif.--Incyte Pharmaceuticals plans to expand its bioinformatics platform into drug discovery and development, as well as clinical trials, with a new emphasis on object-oriented and push technologies, BioInform learned in a recent interview with Scott Clarke, Incyte's chief information officer.
Clarke, who also serves as Incyte's vice-president of informatics, said he sees a vertically integrated future for the company's proprietary databases. "From a software side, we're looking at software products that move more downstream into the whole drug discovery and development process, such as toxicology, pharmacology, and ultimately into clinical trials," he noted.
A six-year-old provider of proprietary genome databases and analytical tools, Incyte is one of the bioinformatics industry's oldest and largest companies. Although it currently has a large client base of database subscribers, it has continued to attempt to differentiate itself from competitors by becoming involved with microarray technology (commonly referred to as gene chips), push technology, and data visualization and mining tools through strategic alliances with other technology developers.
Clarke is one of the architects of the development plan. Two years ago, he was recruited by Incyte CEO Randy Scott to help the firm's products avoid running into what he called "the MIS roadblock" of large pharmaceutical industry information systems integration. Clarke joined Incyte after 10 years with the pharmaceutical company Syntex, prior to its acquisition by Roche, where he headed the research information systems group as CIO. Although "high-throughput sequencing and genomics was a new industry to learn," he said, the industry has similar systems and product development goals, particularly where pharmaceutical information systems are concerned.
Career Comes Full Circle
Now that Incyte is moving toward clinical trial information management--with a genomics angle--Clarke finds his career coming full circle. "I definitely know how to do that," he commented, referring to his Syntex experience with information systems associated with clinical trials. He has also been involved on Incyte's business development side by serving on what he called research committees -- forums in which potential future directions are discussed--with some of the company's partners.
Clarke's group numbers approximately 145 out of a total of 540 Incyte employees. It is responsible for software product development, new technology development, product-oriented bioinformatics development, database quality control, and customer support. Clarke noted that his group's activities have given him a clear perspective on the needs of the field.
"From my perspective as a software support person I can't emphasize this data management issue enough," he told BioInform. "The amount of data that will flood scientists is incredible, and being able to make sense of it and use it will be a huge challenge."
"The ray of hope in this is that we're not the first industry to have to deal with this," he continued, noting that the entire data warehousing and data mining field has already been developed by large financial information and mass marketing organizations that have had to create not only applications but also hardware to deal with enormous volumes of information. "We should be able to ride that wave," Clarke predicted. At the moment, he told BioInform, Incyte's databases are on the order of 10 gigabytes of data, but since other nongenomic databases exist on a scale of 10 terabytes, he was confident Incyte can continue to grow and support databases of that scale.
According to Clarke, much of his approach comes from an information systems, rather than a molecular biology, mode of thinking. In terms of technology development, he said, two high-priority projects will involve breaking down old barriers in information distribution: object relational technology and push technology. Incyte plans to incorporate CORBA, the emerging standard for handling object relational models, in future releases of its databases. In particular, Clarke cited Incyte's collaboration with NetGenics, developer of a CORBA-based viewer for sequence-based information, which should yield a common interface and architecture in the next two years.
Push Technology Will Be a Focus
He also discussed Incyte's recently announced collaboration with push technology provider TIBCO. "The most likely first use of this would be intranet forms of delivery" to commercial pharmaceutical database subscribers, he speculated, where an Incyte database installed within a client's firewall would trigger push transactions that would broadcast customized sequence updates based on user profiles. There would also be value in secure internet delivery of subscription services, Clarke continued, but that raises concerns about data confidentiality. "We're very cognizant and paranoid" about potential security issues involving data transactions over the internet, he emphasized. "The value that we're deriving from our products is in large part due to the proprietary nature of what we have. We will not give that away easily," said Clarke.
Regarding Incyte's current monthly release of updated sequence information, "there is absolutely the temptation," he admitted, to release new data more than once a month using push technology. But the "big issue in terms of the current monthly release cycle is the comparison of new data against all other data" and the need to pull out the "best representations of gene products." Clarke noted that task requires "tons of CPU," hence the monthly schedule. But his group at Incyte is "looking at different ways of minimizing that time by using methods to group the data or using template-based assembly to stabilize the data release process over time so that we could have more frequent releases."
Although the company hasn't started work on a full-blown version of database releases on push, Clarke said Incyte has done a pilot study and "will have some type of push available in the second half of next year."
Meanwhile, Clarke's sights are set on clinical trials as far as Incyte's development plans are concerned. "In five years we will be into the area of providing support in drug development areas and clinical trials," he predicted, explaining that from a drug company perspective, "if you could subset your patient population to either maximize efficacy or minimize adverse effects, it would be an great advantage. Think of how many drugs that actually made it to Phase III testing and failed might in fact be useful for some subset of the population and might provide therapy for that subset," Clarke proposed.
Summing up the fundamental problem at the heart of bioinformatics, Clarke concluded, "This is a data-rich environment, both in the public domain and in the data that we're generating here at Incyte. Data doesn't do any good until it's transformed into information. That's the challenge."