I am an anthropologist and historian of science at Stanford University. I had visited Japan during the summer of 1993 in order to learn about the Japanese Human Genome Project. During this past spring and part of summer, I was very fortunate to be able to return to Japan for a longer visit to the Human Genome Center in Tokyo. I want to thank Professors Matsubara, Sakaki, Kanehisa, Nakamura, and Takagi, as well all the members of the genome informatics laboratory for their generosity in allowing me to spend some time there, to learn about their work, and to learn more about the Japanese Human Genome project. I also thank the members of the Human Genome and Genome Informatics community there for talking to me about their work.

My research interests focuses on bioinformatics in order to understand how biology is changing through the introduction of computer systems. During my 1995 visit to the HGC, I focused on genome informatics in Japan in order to continue my research on bioinformatics and to compare developments in the Japan and the United States.

The introduction of information systems into biology, as you all know, was necessary to help biologists deal with the wealth of information being produced through the various genome sequencing projects. However, the situation appears to be changing. Informatics is becoming an integral part of genomics in ways that go beyond computer scientists helping biologists with data overload. This change is in part due to the desires and visions of computer scientists and in part to the complexities of the vast amounts of data becoming available. Because of these, biologists often cannot even begin to frame their problems in ways that allow them to pose questions to computer scientists.

In the United States, this (proposed) marriage of computer science and biology has led to some friction. Well, perhaps more than some. David Botstein, a molecular geneticist at Stanford's Beckman Center, referred to it as the "two culture" problem. Botstein has argued that the biologist's perception of open, interesting problems might not be interesting to computer scientists or mathematicians. However, those are the problems that need to be addressed and they probably will not be pursued by biologists. He sees a need for computer scientists to be convinced to address these problems.

There are cultural chasms that must be bridged before important problems in biology can be solved. Botstein presented an example where he and a member of his laboratory spent an entire afternoon with an American database expert. They tried to convince the database expert to help them with their database problem. The expert said that the problem was trivial, uninteresting, and that they should "get lost."

Botstein also told several stories where time and effort were wasted because biologists were trying to solve a computer science or mathematics problem without the necessary tools. In one story, a major mathematical biologist solved a problem in one night, after others had run the big computers for days and nights trying to solve the problem.

There are obvious problems that challenge and perhaps might prevent cooperation between "cultures." First, "credit," jobs, and other such things are at stake when working on "other people's problems." "Other people" are clearly people in other disciplines. As Professor Takagi said in his editorial in a recent Genome Informatics Newsletter, this situation can be overcome when computer scientists make biology problems their own. The question is to what extent biologists and computer scientists are able to construct problems of interest to both disciplines.

A potential solution to the problem is to train students to be interdisciplinary, multi-cultured, rather than to try to convince seasoned scientists to work together across disciplinary boundaries. However, both in Japan and in the U.S., interdisciplined students often have difficulties in getting jobs, publications, etc. The disciplines still reign in academia. Until there are a fair number of bioinformatics departments, crossing this particular boundary will remain difficult. (A solution may be to move out of academia into private corporations. This is a matter for another forum.)

After observing and interviewing research teams in the various DOE centers and interviewing individual scientists involved in genome projects here in the U.S., my impressions are that the Japan genome scientists might perhaps be able to avoid some of the problems the American scientists have faced. I say this with some trepidation, since I have mainly focused on genome informatics in Japan and have not spoken with enough biologists to balance my conversations with computer scientists. However, with this primary qualification, I will give you some of my impressions.

There are three ways in which the situation in Japan differs from the U.S. First, the genomics community in Japan is smaller, so there are fewer people to sustain separate communities. For genomics to advance in Japan and to remain "in the game" in the international community, cooperation and collaboration are required. Second and relatedly, the funding situation is different from that in the U.S. Japanese funding conventions, especially under Monbusho, do not allow any one group to become much larger than others. This should also encourage cooperation, since individual or discipline-located empires cannot be built and sustained in this environment. Third, I sensed more willingness on the part of the scientists in Japan to work cooperatively. This might be the impression, rather than the reality, all the scientists wanted to give me. However, if taken at face value, this willingness exists perhaps because of the constraints mentioned in my first two points, because Japanese scientists have learned from examples (successful and unsuccessful) in the U.S. and Europe, or because (as some Japanologists say), there is a greater effort made towards conciliation in Japan. No matter the reason, my impression is that the chances for successful collaborate work between biologists and computer scientists are greater in Japan.

Whether this willingness to collaborate will continue is a question that you all must think about. Collaboration requires respect for each other's "culture," an effort to understand each other's "language" and practices, and a great deal of patience and work. The first two are sometimes difficult to come by in American molecular biology and computer science communities in part because these two disciplines have been relatively well funded in the past twenty years.

Finally, I thank you all for your indulgence. There are many other things I would like to say in this newsletter, but spatial limitations force me to focus on one issue. I hope to continue conversations with you all and to travel to Japan to learn more about your work. Again I want to thank Takagi-sensei and the members of his laboratory for their kindness, generosity, and friendship. I miss our lunches at the osobaya.