LA JOLLA, Calif.--Sequana Therapeutics is evolving from its original focus on gene discovery toward a greater emphasis on functional genomics, BioInform learned in a recent interview with Alan Roter and Andy Watson, who head the company's day-to-day bioinformatics efforts. Watson, Sequana's director of engineering and information systems, also said the company might begin offering the services of its bioinformatics staff to other organizations on a consulting basis.
Founded in 1993, Sequana has been using positional cloning and a proprietary integrated technology platform to search for genes linked to osteoporosis, asthma, type II diabetes, obesity, and schizophrenia. Its initial corporate strategy led to alliances with such pharmaceutical giants as Glaxo Wellcome and Boehringer Ingelheim to handle the development and commercialization of biomedical treatments originating from its gene discoveries. However, in a strategic move consistent with the new interest in functional studies, Sequana announced earlier this year that it will collaborate with Glaxo Wellcome in an arrangement whereby Sequana subsidiary Nemapharm performs functional studies of human genes in the C. elegans system.
Sequana's research efforts rely heavily on bioinformatics. At present, "25-30 percent of Sequana's staff work in bioinformatics," and the numbers could increase, noted Roter, the director of computational biology. The company employs 200.
The foray into functional genomics will increase the complexity of the data flood emerging from Sequana's well-established gene discovery program. Roter said that in the future the bioinformaticists' top challenge will be to keep current systems running and continue to integrate new technologies, such as new informatics algorithms and lab technologies. Watson noted that to support the operational demands of research in functional genomics, his group needs to start developing new supporting lab processes, such as instrumentation, in addition to refining existing methods and developing new software systems for handling the information.
However, he added, "We feel we've done a good job" with developing bioinformatics software on a reliable platform and implementing it for bioinformatics analysis needs. Watson speculated that they might even offer their groups' expertise on a paid basis to organizations outside Sequana in the future.
Until last year, Sequana's bioinformatics operations were run by Carlos Zamudio. When he left (later joining Axiom Technologies), Roter's and Watson's job descriptions were broadened. Watson's group is now responsible for instrumentation and the development of robotics and automation integral to Sequana's high-throughput screening operations, while Roter's group handles analysis of finished data.
Watson reports to Tim Harris, Sequana's senior vice-president of R & D, while Roter runs an autonomous group within the statistical genetics section, headed by Lon Cardon. However, their groups essentially function in tandem as Sequana's analytical bioinformatics unit. Watson's group is in charge of databases, managing the lab aspect of data collection, doing the initial quality control of the data, and developing instrumentation and automation for internal use. Roter's group, composed of two-thirds software engineers and one-third bioinformatics analysts, analyzes data emerging from Sequana's sequencing operations and is in charge of downstream software analysis. It also interfaces between the programming and wet-lab modes of operation.
Watson estimated that support elements, such as computer infrastructure and purchasing and installation of high-powered processors, account for 40 percent of his group's budget, with development of robotics, instrumentation, and automation; software needs; and bench work complementing instrumentation development taking up another 20 percent each. Meanwhile, Roter's group, which is half the size of Watson's, has a large investment in human resources and software for performing downstream genomics analysis, software development, and sequence analysis.
"The biggest challenge is predicting where bioinformatics is going," Roter commented. "Getting a feel for the genome project's output and where that will take us" is crucial. In particular, "the information will need to be put in a form from which people can get actual information and ask biologically interesting questions," he added.
Meanwhile, Watson feels challenged by technological change. "The rapid obsolescence of certain laboratory techniques" is a given, he conceded, adding, "While we recognize that something new will come along every six months, you have to place your stake in the sand at some stage."
Originally trained as a manufacturer engineer at Cambridge, Watson first worked on instrumentation for molecular biology projects at the Medical Research Council in Cambridge. He was part of the original group that founded the Sanger Center, dedicated to sequencing and interpreting genomic data as part of the human genome project. During a two-and-a-half year tenure there, he was responsible for the software side of the organization, working on software control and low-level data analysis. A fortuitous visit to the center by Zamudio brought Watson to Sequana.
Roter also came to the company from a bioinformatics position, but his background was in biology. With a degree from the University of Illinois at Urbana, he went to graduate school at the University of Chicago, where he focused on developmental biology and began picking up programming skills while analyzing data for his thesis. After a postdoc at the Massachu setts Institute of Technology, Roter managed a wet lab for Sandoz Crop Protection, but in his free time developed a molecular biology software tool that later was commercialized. He enjoyed software development so much that he finally chose to leave the laboratory and become a software engineer. "It was a very difficult decision at the time," he recalled, but he took a software developer job with Applied Biosystems. Five years later he joined Sequana.
Watson acknowledged there was a mental leap to be made, moving from an analytical background to molecular biology, but said he found it "quite easy" to learn enough about molecular biology and sequencing to manage the wet lab people in his group. Roter added that as a biologist-turned-informatician, he primarily learned software developer skills hands-on, although he did take a couple of classes in programming along the way.
In addition to moving from engineering and biology, respectively, into bioinformatics, both Watson and Roter have moved into another new field: management. An immediate challenge has been staffing; Watson said one of the hardest aspects of his job is finding and keeping people with a talent for bioinformatics development. "It's important we provide a working environment that people feel is the best for them," he observed.
A vision of the future
Asked what he wants to see happen in bioinformatics over the next few years, Watson replied, "I'd like to see the development of databases that use the HGP information and return results without the need for human intervention." Today, many experiments give ambiguous results that can involve a considerable amount of manual interpretation. For the millions of experiments that make use of human genome sequence data, Watson believes it will be essential to develop assays and instruments that present unambiguous data that can be automatically entered into a database and used to answer biological questions.
Roter, on the other hand, would like to have software access to public databases become not only better standardized, but more powerful. The current de facto standard is Web-based, he noted, which allows easier access to information but reduces the power of software tools that can access, query, and analyze results of queries. An alternative technology, Common Object Request Broker Architecture (CORBA), is object-oriented and allows automated software tools to ask more sophisticated questions, but whether it becomes widely accepted depends on the investment the bioinformatics community is willing to make, he said.
As for the future, in 10 years, said Roter, "I'd like to be the CEO of Sequana or a genomics company like Sequana. I have a lot of confidence in genomics and want to be a player." Watson had a similar vision, of "starting a small company" in bioinformatics, perhaps dealing with instrumentation. "Bioinformatics is the future for genomics," Roter concluded.