SEATTLE--At 60 years old, Leroy Hood, chairman of the Department of Molecular Biotechnology at the University of Washington School of Medicine, is something of an old-timer in the young field of genomics. Yet, from the formation of a new institute to advance molecular and genetic medicine to the identification of a gene believed to cause an inherited form of prostate cancer, he seems to have a hand in all the cutting-edge developments emerging from his institution lately.
Hood, who holds an MD from the Johns Hopkins School of Medicine and a PhD in biochemistry from the California Institute of Technology, recently collaborated with the Fred Hutchinson Cancer Research Center here to pinpoint chromosome 1 as the location of a prostate-cancer-causing gene, a finding that could lead to new treatments and diagnostic tests for the inherited cancer. In conjunction with the Human Genome Project, Hood has also applied his laboratory's expertise in large-scale DNA mapping and sequencing to the analysis of the human and mouse T-cell receptor loci. And, he is a founder of Rosetta Inpharmatics, a privately held bioinformatics company in nearby Kirkland.
Hood met with BioInform recently to discuss how he sees the role of bioinformatics evolving as genetic medicine advances, and to share his thoughts about how bioinformatics vendors can succeed.
BioInform: Are bioinformatics companies on track in terms of providing software and tools to help researchers like you succeed?
Hood: What I think they are doing right is many of them are integrating tools developed by academics and putting them into their packages. An example is the software that someone such as Phil Green here has developed. I think what the companies have done is much more software engineering than creative new developments. I don't know that any of the bioinformatics companies out there have succeeded in establishing an identity that makes everyone say, gee, they have a product that I really like. There is a lot of similarity. There isn't product delineation and definition. A lot of the bioinformatics companies think they can do the software integration independently of effective groups that are doing the biology and the sequencing. That's a big mistake.
BioInform: What will they have to do to succeed in this field?
Hood: They need much more effective partnerships with groups that are doing the experiments so that there is that continual testing and refinement. For business, the really key thing is what your customers need. To give them what they need, there has to be a real integration of software with instrumentation.
BioInform: A recently published study says that, despite high market value, segments of the bioinformatics market could become saturated. What are your thoughts?
Hood: I think it's pretty unlikely that the bioinformatics market will get saturated because it's evolving at such a rapid rate. Two years ahead of time there will be a whole series of new demands we haven't even thought of that software companies will really have to respond to. The ones that are going to survive are those that are agile, adaptive, and very effective in creating partnerships, and then picking up these new opportunities quickly.
BioInform: Where do you see the genomics field heading, based on the research that has been accomplished to date?
Hood: Essentially what we've done with the genome project is create the periodic table of life that is totally equivalent to the periodic table of the elements. In the 19th century those relationships revolutionized how we did chemistry. In the 21st century they are going to revolutionize how we do biology.
The genome project is different from any kind of science that has ever been done. It's what I call discovery science and it's a very, very different kind of science. It's a science that essentially creates infrastructure. Now, in the context of doing that discovery-driven science, we've created what I call global tools; that is, high-throughput analytic tools for sequencing DNA, for genetically mapping DNA, for doing DNA array studies to look at expressed genes, and so forth. The idea of global tools is to look at many elements at a time and analyze them very rapidly.
The biology of the future is studying systems. So the biggest heritage of the human genome project is that it has given us the tools, the periodic table of life, the instrumentation, the approach to software that we need to do systems biology. That's going to be the big heritage in five years, in 10 years, and in 20 years.
BioInform: Once the human genome is mapped, where will your research go next?
Hood: We'll start studying systems. How does autoimmunity work? How does the immune system work? We'll start looking at a systems approach to disease, such as prostate cancer.
BioInform: What does the best bioinformatics software offer?
Hood: Let me tell you what it doesn't offer. There are fundamentally three types of biological information. There's the digital information contained in our chromosomes and genes--that's what genomics is all about. There's the three-dimensional information contained in proteins and other molecular structures like carbohydrates. And then there is the four-dimensional information contained in biological systems and networks.
I would argue that our software doesn't know how to handle the systems information at all. We are at the very beginning of being able to handle protein information. But there isn't any company out there that has software that can handle all three of those types of information. The challenge is integrating all this information, and converting DNA sequences into knowledge. What bioinformatics companies today can handle is a little bit of the information. They don't have any clue how to convert it into knowledge.
BioInform: Any thoughts on how bioinformatics tools developers can get to that point?
Hood: It's an enormous integration of the biology with the software. Most of the companies won't even begin to get any place until they have partners who can do this integrating much better.
--Amy E. Nevala