intramural scientists discussed these issues, and they felt fairly strongly that the secrecy issues and the reach-through provisions were sufficiently troubling for the intramural program to choose not to make use of any agreements with HGS.
We acknowledge that a company like HGS needs to make some profit on its investment in ESTs. Here is one place where self-interest obviously creates a difference of opinion about what should be done. It is interesting to contrast how ESTs are offered to individual investigators with how restriction enzymes are offered. Both are products of companies. If we buy a restriction enzyme for $50 or $250, there might or might not be a patent on it. We are pleased to pay the asking price because it is an efficient way for us to get a useful research tool. No one attempts to seize downstream rights on anything that we do with that restriction enzyme. In contrast, the reach-through provision attached to the use of ESTs creates a very serious problem for us, and that (tied in with the secrecy arrangements) has discouraged us from pursuing the EST databank that is made available through level II agreements.
What should be patented, and what should be placed in the public domain? I see this as the central question. Approaches to such questions are influenced by three major factors: self-interest, the investment of the public in government-supported research, and the law. How do we determine whether our answers to the questions are correct? With respect to what is legal, one must simply wait for decisions from either the judiciary or the Patent and Trademark Office. The more interesting issue is how we judge which decisions to make with respect to the goals of NIH and the interests of the public. It is sometimes easier for a company to judge whether it has made the right decision. Did it make a profit? Did it get more investment capital? A university, in some cases, might try to weigh the effectiveness of what it does on the basis of whether there are royalty returns. But, as Lita Nelsen pointed out, there are many other reasons why a university would want to get involved in technology transfer and intellectual property protection—reasons that are much harder to measure.
From NIH's perspective, how do we measure whether the scientific community is more productive as a result of decisions that we have made? How do we measure whether applications of knowledge occur more efficiently? These are difficult questions that we would like to know the answers to, but we do not.
One way to look at this is to follow the outcome of the use of sequences that are in the HGS-TIGR database (TIGR, The Institute for Genome Research is a nonprofit partner of HGS), as opposed to sequences that have been put into the public database as a result of the activities of Merck. Will sequences that investigators obtain only by going through level II agreements with HGS produce more benefits than those studied by academic investigators and obtained free of any attachments from Genbank after being sequenced by Washington University and paid for by Merck? This would be a useful experiment.
One of the things that we learn from these discussions is that every example of a research tool that breeds contention has its own characteristics and its own