8
Conclusions and Recommendations

The foregoing case studies by no means exhaust the list of successful efforts to share biomedical data, materials and facilities with the scientific community as a whole, but the common themes that emerged in discussion of this diverse group of cases encourage the committee to believe that they are representative of the equally successful ventures not considered because of constraints on the committee's time, energy, and funding. These common themes demonstrate some of the necessary ingredients for successful resource sharing, but also surface issues or problems that require further study.

Features of Successful Resource Sharing

Strong Scientific Leadership in Agencies and the Research Community

Essential ingredients in successful resource sharing are the leadership of program managers in government agencies who identify opportunities and support them; the leadership of senior scientists who establish the norm for the scientific community by example and commitment to sharing resources; the leadership of scientists who direct existing shared resources to provide quality services at moderate costs; and the commitment of scientific institutions such as universities and professional societies that develop policies to facilitate and enforce resource sharing. The Arabidopsis thaliana genome project's remarkable communal spirit and international character have made it a successful model for scientific cooperation and sharing of research resources.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 71
--> 8 Conclusions and Recommendations The foregoing case studies by no means exhaust the list of successful efforts to share biomedical data, materials and facilities with the scientific community as a whole, but the common themes that emerged in discussion of this diverse group of cases encourage the committee to believe that they are representative of the equally successful ventures not considered because of constraints on the committee's time, energy, and funding. These common themes demonstrate some of the necessary ingredients for successful resource sharing, but also surface issues or problems that require further study. Features of Successful Resource Sharing Strong Scientific Leadership in Agencies and the Research Community Essential ingredients in successful resource sharing are the leadership of program managers in government agencies who identify opportunities and support them; the leadership of senior scientists who establish the norm for the scientific community by example and commitment to sharing resources; the leadership of scientists who direct existing shared resources to provide quality services at moderate costs; and the commitment of scientific institutions such as universities and professional societies that develop policies to facilitate and enforce resource sharing. The Arabidopsis thaliana genome project's remarkable communal spirit and international character have made it a successful model for scientific cooperation and sharing of research resources.

OCR for page 71
--> This project began when program managers in government agencies, recognizing that work on mapping and sequencing the genome of Arabidopsis was accelerating, convened an international series of workshops of leading scientists to devise a long-range plan. The continued commitment of these senior scientists to widespread sharing of information and materials, and the peer pressure and aggressive solicitation of stocks of mutant strains to be made available through distribution centers, have contributed to the almost universal sharing of materials in this community. Similarly the strong leadership of the 22 societies that provide oversight for the American Type Culture Collection (ATCC), and the strong scientific leadership and management of The Jackson Laboratories (TJL) are strengths of these successful repositories and distributors of resources. A most remarkable example is presented by the Human Genome Center of the Lawrence Livermore National Laboratory (LLNL), which, by default, has become a major supplier of material resources to the scientific community, without being supported for this function. The extent to which it has provided the leadership and the actual materials that have permitted widespread sharing of genetic materials and information and the forging of important collaborations is remarkable. LLNL has protected the use of this important resource for the research community. Many of the important institutions in science have an ongoing responsibility to foster a culture of sharing and to continue to advocate for policies that assist the process. Professional societies and journal editors can support sharing of resources by developing appropriate policies guiding publications and responsibilities for making data available after publication. The Journal of Biological Chemistry, for example, has such a policy: ''Authors of papers published in the journal are obligated to honor any reasonable request by qualified investigators for unique propagative materials such as cell lines, hybridomas, and DNA clones that are described in the paper.'' Plans are under way to modify the phraseology to restrict the obligation to investigators who want to use the strain for noncommercial purposes and to include computer programs in the materials that have to be shared. In addition, after considerable debate, the policy was established that authors publishing crystallographic data must submit the details, coordinates, and related data to the Protein Data Bank at Brookhaven before publication. The appropriate accession number must be inserted into the manuscript; in a similar way, nucleotide sequences must be submitted to Genbank or a similar database, and the accession number must be inserted into the manuscript.

OCR for page 71
--> Adequate Core Funding The committee observed that an essential ingredient for successful shared facilities or repositories was adequate funding of the core functions. In many cases there is a patchwork of funding from a number of different funding agencies, industry, and grants to support research or further development of the resource, as well as user fees. Sometimes the different streams of dollars may not be available to support the core administration and quality control necessary for resource sharing. This is inefficient and requires much effort on the part of the staff to write numerous proposals to different agencies. For example, at ATCC, decreasing core support is a cause for concern that has forced management to raise costs to purchasers to undesirable levels. The MacCHESS (Macromolecular Crystallography Resource at the Cornell High-Energy Synchrotron Source) case story is an excellent example of coordinated agency and industry support. The National Institutes of Health (NIH) is able to piggyback on the support provided by Department of Energy (DOE) and the National Science Foundation (NSF) to open these facilities for use by the biomedical community. The DOE scientific facilities initiative of FY 1996 provided these facilities with increased operational funding to ensure full-time operations and effective running. The seven regional primate research centers established by specific legislation during the 1960s and funded through the National Center for Research Resources are additional excellent examples of shared resources that have stable core funding. Marketing and Advertising Advertising, marketing, and general knowledge of the availability of a resource are essential to widespread access; many resources are not shared simply because their existence is not known to scientists who require them. All of the case examples studied in this workshop have a variety of mechanisms for alerting the research community about the availability and costs of their resources. From a marketing point of view, for example, ATCC has a very heterogeneous user group, supplying materials to the clinical, industrial research, university, and government markets, and it reaches these groups through a variety of printed media, electronic media, and workshops. The Jackson Laboratory provides a variety of price lists, lists of stocks with genetic information, data sheets on individual strains, newsletters, and a handbook on doing research in mice. Most of these are also available electronically. A unique resource is the Primate Information Clearinghouse set up by the Washington Regional Primate Research Center (WRPRC) in 1977. This is an international effort to list available primates and researchers desiring primates, as well as to provide literature reviews and other information such as annual

OCR for page 71
--> reports, and regulations. The goal of this very extensive effort is to ensure that every animal is utilized to its fullest extent in research to minimize waste or needless use of animals. Clear Guidelines about Ownership and Access The cases reviewed at the workshop demonstrated the value of clear guidelines concerning access and ownership, although these differ depending on the resource. No single approach can accommodate the different uses or needs. Project planning should include guidelines for sharing—under what circumstances and with whom data and materials will be shared. This is an essential ingredient in preventing later misunderstandings and problems. There is increasing desire to commercialize and realize the economic benefits of biomedical research, which makes this an especially important and changing feature of shared resources. At ATCC, special collections are being developed with restricted access, and new policies have been formulated to clarify ownership at the time of deposit, with a heavy emphasis on donation to ATCC with no restrictions. In the case of Arabidopsis, the stock centers and databases do not permit restrictions on materials, and strong scientific leadership and peer pressure serve to make these materials and the data freely available to the research community. The Jackson Laboratory provides another example of a resource that has developed explicit policies on ownership and access, and is resisting licensing agreements or agreements that give reach-through rights to commercial entities. The Human Genome Center at LLNL similarly has developed policies to address access to information and materials it distributes in order to protect access for the rest of the research community. For example, LLNL has no bar to commercial use of individual clones but does bar commercial use of whole chromosome-specific libraries. User Fees One important source of funding for shared resources can be user fees. These charges help to subsidize the core operations and maintenance of those research resources that are not currently commercially viable. In addition, at both TJL and ATCC, fees from sales (mice at The Jackson Laboratory and cultures and cell lines at ATCC) help defray the costs of functions such as authentication and quality control, which are essential, if invisible, elements of first-class science.

OCR for page 71
--> Clear Policies for Retaining and Discarding Data and Material There are substantial costs associated with sharing of materials and data. Policies for the disposition of materials and information that are no longer of value will be increasingly important as the body of resources that need to be shared continues to increase more rapidly than the funding available to support them. At The Jackson Laboratory, for example, if a mouse strain is not requested for six months, the strain is stored through cryopreservation, but live colonies are no longer maintained. Prioritizing which resources to support and which not to support will be increasingly important. When the growth of different induced genetic mouse strains recently outpaced the capacity of TJL to produce these for the larger research community, the laboratory established an advisory committee to decide priorities as well as seek additional funding from government agencies. Quality Control A critical attribute of a shared resource is that the distributed resource be what it is purported to be. Similarly, mechanisms to ensure the highest-quality research at limited-access resources such as a synchrotron are essential to their ongoing success. The Jackson Laboratory is an excellent example of intensive quality control. First, all mice obtained from the facility are of known health status and genetic quality. Any mouse released by TJL is genetically defined so that individuals who obtain mice will continue to receive genetically identical animals. Strict distribution rules protect and ensure the quality of TJL animals. Scientists are asked to return for new breeders after 10 generations and to limit distribution to their own institution. ATCC also has a long history of providing well-defined and reliable cultures to the research community. MacCHESS, which represents a saturable resource and thus a different dimension of quality issues, has developed an excellent proposal process and peer review system to facilitate access to the synchrotron and to ensure that only the highest-quality research is conducted at the facility. Well-Defined Policies for Function of Research and Service at the Facility The balance between service and research by staff is a fundamental question to be considered by all centralized facilities designed to be resource centers for the scientific community. A shared resource is greatly enhanced by the presence of an excellent scientific staff that is conducting research to

OCR for page 71
--> improve the resource and can ensure the quality of the materials. Strong scientists at the resource can also collaborate with and expand the ability of outside scientists to contribute to new knowledge. All of the case studies have strong scientific staff that conduct research to develop the resources, are critical to quality control, and also collaborate widely with outside scientists. Sophisticated Information Retrieval and Transfer Systems Rapid exchange of information and widespread access to data are greatly facilitated by sophisticated information retrieval and transfer systems. Rapidly evolving information systems are transforming the way research is conducted and disseminated. A decade ago, a paper that reported an extensive body of DNA sequence data was a landmark. Now such data cannot be published in scientific journals at all but are deposited in data banks. In the case of the Arabidopsis community, a sophisticated set of databases and links among them facilitates reaching the entire research community on an ongoing and almost instantaneous fashion. As soon as genes are sequenced in Chris Somerville's laboratory, for example, the data are sent directly to the University of Minnesota, where the initial analysis takes place. Similarly, information generated by LLNL staff and collaborators goes into the genome database funded by DOE, where the rest of the scientific community has ready access to the information. Issues and Problems No meaningful argument can be made against the sharing of scientific resources. No convincing example exists where sharing has had preponderantly damaging or deleterious effects. Sharing almost always results in a total cost reduction, allowing existing research dollars to support a larger total research effort. Sharing has other side benefits including the rapid diffusion of new techniques or methods throughout the scientific user community and, quite often, the catalysis of scientific collaborations based directly or inadvertently on the sharing experience. The issue is, then, not whether there should be sharing, but how to optimize it. The case studies, although providing many good examples of "best practices," also provided the committee with a wealth of unresolved issues and emerging problems that any future sharing effort will have to address.

OCR for page 71
--> One Uniform Policy on Resource Sharing is Not Possible The problems of sharing resources are diverse. Solutions therefore will be similarly diverse. There are differences in the resources to be shared, the needs of stakeholders, and the distribution of resources that stakeholders command. In gathering the material for this report, the committee has dealt with the sharing of data, materials (including experimental subjects), and equipment. It is clear that the optimal procedures for sharing these three classes will differ in most cases. With data, the incremental cost of sharing or wide distribution may be negligible. Thus, sharing as broadly as possible should be the community norm. The amount of regulation or review needed to ensure standards and effectiveness in such sharing can be minimal. Successful examples of such sharing include the nucleic acid, protein sequence, and similar databases (e.g., Genbank, DNA Database of Japan, Genome Science Data Base, SwissProt, Protein Data Base, Genome Data Base), which operate as worldwide consortia with free access to all users. Materials (or experimental subjects) fall into two classes. Some materials are renewable. Examples are clones, polymerase chain reaction (PCR) primers, strains, and most transgenic animals. Here broad sharing is to be encouraged because it is cost-effective. However, the incremental costs of sharing are significant, and mechanisms to distribute these costs have to be developed and optimized. It seems advantageous to avoid a situation in which no costs accrue to the end user and there is no incentive to be frugal or cautious in requesting materials that may not be essential. Other materials are not renewable, such as some clinical samples, unamplified libraries, extraterrestrial samples, deep sea or deep drilling cores, and fossils. How these samples are treated for possible sharing will have to be dealt with largely on a case-by-case basis. The overall guiding principle in such decisions should be scientific merit and the acquisition of information of interest to the scientific community at large. Equipment, unlike data and most samples, is saturable. In addition to an incremental use cost, the total amount of available access is limited. Some animal resources are also saturable. For example, the number of animals that a primate center can produce and maintain is certainly finite. Here, a proper balance needs to be struck between acknowledging or rewarding those who had the foresight to construct, acquire, or fund such equipment and the desire to see equipment (or animals) be available for use by the highest-quality scientific projects, wherever they arise. Some facilities, such as synchrotrons, are best viewed on a worldwide basis. Others will be best managed on a national, regional, local, or institutional basis. A general guiding rule that seems applicable to most cases is dividing available time so that those who are responsible for the resource have significant privileged access, but the remaining access to the resource should be competitively available to all users. External use should be judged by scientific quality and by the need for access

OCR for page 71
--> to the unique resources. Under ordinary circumstances, whether or not an external project is directly competitive with one already ongoing at or planned for the shared facility should not be considered in making this assessment. Occasionally, when a project is extraordinarily taxing in terms of the time or staff available at a resource, competitive projects should be discouraged. Here, the potential competitors should, if at all possible, be encouraged to work as a team. If this fails, first come first served seems like the only simple system to resolve the conflict. Incentives and Rewards for Resource Sharing are Not Fully Developed The current systems for rewarding academicians or employees in industry do not encourage sharing but rather focus on individual achievements. There are no simple answers to questions such as the following: how much "credit" should an individual receive for providing transgenic animals or research reagents to colleagues, and for what period of time? How should the collaborative contributions of individuals scientists to research projects be evaluated? Sharing Requires Incentives, Not Disincentives For academic scientists, incentives are citations or other credit for use of samples made available; another incentive is having the costs of making these samples available covered by the recipient, a third party, or one's grant. Incentives also need to be offered for those who make raw data available over the World Wide Web, since some remote reprocessing of raw data will inevitably be quite valuable. A foreseeable shift in emphasis toward more theoretical or computation biology means that the impact of sharing data that is not normally in public databases must be addressed in a timely fashion. Provisions for sharing data, materials, and equipment should be built into research proposals, and the sharing activities should be included as part of the progress report when grants are being considered for renewal. For all sharing of materials, data, and equipment, there is a temporal threshold after which the individual investigator should be removed from the loop (i.e., although soon after discovery, an investigator might reasonably demand coauthorship from others using his or her resource, after some period only an acknowledgment is appropriate). The willingness of scientists to participate in the Arabidopsis project was enhanced by the scientific credit they received for participating as well as the peer pressure exerted upon those who were less enthusiastic participants.

OCR for page 71
--> Likewise, the major incentive to an investigator to contribute animals to TJL is that he or she receives considerable scientific credit and also frees him- or herself from maintaining a colony to supply peers with animals. Similarly, it frees him or her from the attendant issues of shipping, monitoring, advising, et cetera. The disincentives are that it increases the competitiveness of scientific peers and is an expense to the contributor—although this may be charged to grant support or may be supported by the contributor's institution. A different type of disincentive occurred when there was an exponential growth in requests for materials from LLNL, which was inadequately funded to support these requests and received little or no scientific credit for providing these resources. The Importance of Material and Data Assets Changes Over Time A key clone at the early stages of an investigation may be worth trading only in an actual scientific collaboration. Later, the clone may be freely available in a public repository or distributed upon request. Finally, the clone may become archaic: it should not be kept or distributed; public repositories should deaccession it. Technologies and Needs are Evolving Very Rapidly Any system put into place must have sufficient flexibility to evolve as well. New Definitions of "Publication" May Have to Evolve to Keep Pace with the New Electronic Information Systems It is remarkable that over the past two decades at least a millionfold increase in the power of computer hardware and software has occurred without any significant impact on the way credit is awarded in the university research community for work performed and reported. Should the inclusions of methods, sequences, or other data in readily accessible databases have some relative merit compared to scientific articles and book chapters? Ways of providing credit to institutional shared resources must be found, or support for the scientific mission of these core activities—which benefit many—will be endangered.

OCR for page 71
--> Methods for Enforcing Existing Policies on Sharing are Inadequate Although some policies already exist mandating sharing, most notably that of the Public Health Service in regard to products of research with public monies, the enforcement of these policies is inadequate. It is possible that better rewards for sharing will make failure to share sufficiently unattractive that no explicit sanction is necessary, but until that time it seems only logical to discourage noncompliance at the same time as we reward compliance. Should universities be the main point of enforcement? To what extent should government funding agencies take a role in enforcing sharing? How should the willingness to share impact funding? The role of universities and professional organizations in encouraging and facilitating sharing was prominent in workshop presentations and discussions. Actions against scientists who fail to share, however, are rare. To a very small extent, NIH has required sharing or withheld funding (especially for structural data). It was unclear whether NSF has taken the same position. The policy stated, but also not rigorously enforced, by many journals that a published clone or other renewable sample should be available publicly is a sound policy. The issues yet to be resolved are the actual mechanism of enforcement and how the costs involved should be paid. There are Many Private and Public Stakeholders in any Major Resource Sharing Attempt, Often With Conflicting Goals The boundary between private- and public-sector activities that impact on shared resources is complex and raises issues that will need to be monitored carefully. The National Center for Biotechnology Information (NCBI) provides an interesting example in which, as a compromise, a federally funded public database will make some software publicly available, but the provision of commercial quality supported software is left largely to the private sector. DNA synthesis and DNA sequencing are two other areas in which the needs of the community and the activities of the private sector will have to be balanced. It does appear that economies of scale will dictate that some such activities are better provided as private-sector services as long as actual costs to the users do not inhibit research. In WRPRC, the ownership of the monkeys is retained by the institution, but use of the animals resides with the scientists after appropriate peer review. When internal review committees for saturable resources such as nonhuman primates or synchrotrons exist, however, concerns about conflict of interest

OCR for page 71
--> between internal scientists competing for the same resources must be closely monitored. Resources may often go to those who possess the most money to pay for them or who have the freedom to profit from them. This may place equally or more creative scientists, who are less well off financially at a disadvantage. For example, the costs of some mouse strains at TJL is driven up when for profit groups "cherry-pick;" for example, they undercut sales of popular mouse strains by TJL by marketing only the most financially viable animals and ignoring less commercially appealing strains. TJL maintains the latter as a service to the research community, using revenues from the former to offset the costs of maintaining the colonies. The perception that scientific data and research materials (animals, reagents, etc.) have potential commercial value frequently causes universities to be even more reluctant than individual scientists with respect to sharing. The relationship between intellectual property issues and sharing is a complex one. It rests on ambiguities in current issues of credit and ownership that go beyond the additional constraints imposed by sharing. These issues are badly in need of clarification and resolution. One example is the status of the research exemption from licensing for university-based investigations in a climate where universities are required by law to protect intellectual property that is potentially valuable commercially. The current status of PCR patents is one area ripe for such investigation. It must also be recognized that different cultures regarding sharing may exist within academia, or industry and among individuals scientists irrespective of their place of employment. Industry is generally thought to focus heavily on retaining intellectual property rights by stringent enforcement of confidentiality and material transfer agreements. Efforts to protect the long-range interests of stockholders may involve demanding far-reaching agreements that make ownership claims on future inventions related to the material or technologies industry produces. The activities of Bristol Myers and the human immunodeficiency virus strain HIV-2287 are an example. Other companies have demonstrated a more thoughtful, long-range concept of value. The government and the scientific community should seek ways to foster this more enlightened attitude. Who Pays and What Do They Pay for? The issues of quality control and quality assurance for shared samples or sample repositories are of major concern. Sharing of individual reagents even within single laboratories is often compromised by concerns about improper prior sample handling. Both TJL and ATCC have resolved these problems by characterizing the animals and materials they provide. Mice from TJL are of

OCR for page 71
--> known genetic background, which is constantly monitored, and have been caesarean rederived to eliminate diseases that will affect research results. JAX mice may be more expensive than those from other suppliers that do not provide the same quality. Commercial competitors willing to employ less stringent measures on a smaller selection of resources can and do offer apparently similar products at cut-rate prices. High-quality research depends on high-quality materials, and the scientific community will have to recognize that it must pay for quality control, through subsidy if not through user fees. Similar issues regarding quality control may exist for shared data. How well are the data validated? No simple universally applicable answer emerges, but a combination of improved analytical tools for quality assessment and user education about proper sample handling methods will help to reduce costs incurred by wasted or contaminated samples considerably. It is worth noting that for chemicals or reagents, where any kind of hazard is involved, the cost of disposal often dwarfs the cost of acquisition. This argues strongly for virtual supplies, stockrooms, or repositories where samples are not created or subdivided until they are needed. Such a scheme will work only with an extremely efficient distribution system. The use of electronic ordering, inventorying, and purchasing will become the norm, and this should help encourage efficiency. A key ingredient to quality control is the funding for key administrative and support personnel who carry out this essential, but relatively low-profile, activity. Regulatory Requirements and Documentation Can Be Unnecessarily Complex and Burdensome Regulations promulgated by government agencies affect shared resources disproportionately. The regulatory burden on ATCC for shipping many samples is necessarily greater than that on an individual who ships an occasional sample. Some regulations governing animal care and shipping by the various municipal, state, and federal agencies are conflicting. Regardless of their scientific basis, the costs of complying with these regulations and the extra documentation required by them add burdens to the individual scientist, his or her institution, and the shared resource. Among the underlying reasons for the centralization of primate center programs, for example, was the desire to increase animal welfare and decrease cumulative regulatory costs; despite this, regulations and requirements for documentation for the use of animals or animal tissues continue to increase exponentially. A second issue, only tangentially addressed in the workshop but potentially stifling to some sorts of clinical research, is the increasing

OCR for page 71
--> regulatory activity regarding the use of human tissue and tissue products in research. Who ''owns'' these materials and what sorts of informed consent must be obtained before they are used or reused? Education of Scientists Covers Neither the Ethos of Sharing Nor Intellectual and Tangible Property Issues There is a significant gap in leadership in the training of scientists with regard to the issues of intellectual and tangible property: What constitutes intellectual property? When and how can (or should) patents be used to protect individuals and institutions? During training, there is no formal emphasis on the merits of sharing or the benefits of collaborations, and in an increasingly competitive atmosphere where resources are limited, the benefits of sharing may be unappreciated. Resource Sharing Can Have National and International Implications What are the consequences on the U.S. position in international trade of complete government funding of national culture collections (e.g., in Germany and Japan)? What guarantees are there of future access by U.S. companies, and individual scientists? In various countries the relationships between business and government differ, and the support for core shared facilities that benefit business often derives from the government. How will such national authorities interact with countries such as the United States that are in turn providing resources to them? What benefit is there to various governments to duplicating databases and collections? How will countries that have different interpretations of intellectual property treat scientists from other countries? What protection can these scientists anticipate? In underdeveloped countries will the desire to protect what are perceived as national resources, such as plants or animals, impede the free movement of materials and animals? Wherever resources are saturable or irreplaceable, all efforts should concentrate on viewing the scientific utility of such resources from a worldwide perspective. Procedures should be developed for worldwide review of competing applications for limited resources or facilities. Synchrotron x-ray sources are one area ready for the early implementation and evaluation of such procedures. Ecological and environmental samples, and strain collections are other areas in which a worldwide perspective is absolutely essential. The United States is in a strong position to catalyze such global efforts because, today, it has a major position in shared scientific data—a valuable resource that is already made available on a worldwide basis. The National Research

OCR for page 71
--> Council is in an excellent position to work to realize these goals by networking with other academies and relevant government agencies worldwide. There is a Gap in Leadership Sharing of research resources lacks high-profile leadership (for example, the president of a major scientific society or the president of the National Academy of Sciences). Universities, government agencies, and industry have failed to focus the scientific community. Partnerships in Sharing Resources May Be Unequal The issue of fairness in access and opportunities to utilize shared resources is ongoing because there are typically inequities between those seeking access to saturable or costly resources. For example, graduate students or junior faculty may seek resources from large companies or senior investigators but have little to offer by way of a collaboration, whereas a more senior investigator seeking the identical resource may be perceived as an attractive collaborator. Monopolies Can Be Good or Bad Federal funding policies typically require competition for funds, but in some cases this may be an artifice that is unwarranted. Although a competitive renewal of primate centers might elicit some creative new ideas, it seems less certain that requiring individuals to submit proposals that will compete in setting up stock centers and services for Arabidopsis is serving either science or taxpayers well. The goal should be to identify the most cost-effective methods and highest quality resources for the scientific community. Recommendations This study is exploratory in nature rather than definitive. The committee was not asked to provide solutions so much as identify present and future obstacles and point out directions for followup in more definite studies by a similar committee or others. The committee believes the Academies are in a unique position to provide leadership and bear some responsibility for the culture and ethos of sharing. As a result the committee recommends study or

OCR for page 71
--> further work to address a number of the problems and issues raised in the workshop summarized in the previous section. Administrators of research institutions, grant administrators, scientists, and industry representatives should meet to develop policies to foster sharing of resources. These policies should explicitly address the following: • Sources of reliable funding for provision of materials and services to the research community. A portion of the costs of sharing should certainly be borne by the requestor of the material or service. In some cases such user fees might cover the entire expense incurred by the provider, but in other cases setting fees at that level would effectively preclude sharing with much of the nonprofit research community. Several of the case studies instead subsidize the sharing of materials, equipment, or services from funds the primary purpose of which is not sharing, just as individual scientists use research grant funds to provide materials to colleagues. Funding agencies should consider more straightforward mechanisms by which facilities might be reimbursed for the full costs of sharing with the rest of the scientific community. One possibility might be peer-reviewed distribution contracts providing reimbursement of costs not covered by user fees. The duration of such contracts should be long enough so that grant writing is not a major activity of the facility, and the need for competitive bidding not so great as to preclude awards to a single competent facility. • Training and education regarding the ethos and the value of sharing and related intellectual property issues, including the merit of patents and licensing Education in these matters needs to begin early in graduate training and should parallel educational offerings in the area of scientific integrity. As with scientific integrity, education in scientific sharing needs to be strongly reinforced by an environment within the institution that demonstrates willingness to share and the benefits to be derived from such behavior. Ergo, university administrators as well as scientists need this education and training. • Rewards and incentives for researchers who share resources To foster an environment that can serve as a model for the appropriate education of graduate students and induce researchers to share, it is necessary

OCR for page 71
--> to develop incentives for those who do share. This means that there must be recognition in terms of academic credit, promotion, and salary for those who share. As a concrete example, acknowledgment for having provided a critical reagent in a significant paper should carry a proportional benefit relative to having been an author of such a paper. In the same manner, funding agencies could make resources available on the basis of such sharing, perhaps by requiring applicants' biographical sketches to include such items as provision of resources to other scientists or repositories and memberships on shared resource steering committees. Deans, department chairs, and other university administrators might then come to view membership on such committees as a prestigious appointment similar to membership in a study section. Grants might also provide additional funds to cover expenses incurred in sharing materials with other scientists. • Mechanisms for enforcing agreed-upon resource sharing policies within and across institutions The funding agencies have a clear stake in promoting the optimal use of research resources, and in some cases already have articulated clear policies mandating sharing. They are however ill-equipped to investigate allegations of violations, and have as a penalty for noncompliance only the all-or-none revocation of funding. Because the local research institution controls the employment, reimbursement, academic rank, and space available to the researcher, it is potentially the most effective enforcer and in the best position to determine the extent of enforcement required. Research institutions, however, as well as the scientific societies and journals that provide scientists with recognition, do not have the same obvious stake in sharing as the funding agencies. The funding agencies may therefore have to begin this task by arranging a stake in sharing for these institutions. The resulting cooperation would have a synergistic effect regardless of the extent to which both institutions and funding agencies should encourage or insist on sharing. • Role of the technology transfer office in facilitating resource sharing In several instances during the workshop the statement was made that the institutional technology transfer office was often more of a hindrance to sharing than the individual investigator. Clearly, the technology transfer office has the obligation of protecting the researcher and the institution with regard to intellectual and tangible property; however, there has already been significant progress in the development of uniform material transfer agreements between not-for-profit institutions. This and other such mechanisms can foster

OCR for page 71
--> sharing and should be aggressively developed and used. Similarly, the strong "advertising" programs of many of the case studies suggests an important proactive role for technology transfer offices, publicizing resources at their institutions available not simply to for-profit partners but to scientists at other academic institutions as well. • Current National Institutes of Health guidelines governing university-industry relationships The current NIH guidelines governing relationships between universities and industry encourage institutional patenting of NIH-sponsored research results and licensing to industry. Thus, the question arises of the extent to which reagents and results originally dependent on public support should be shared versus the initial period of confidentiality sometimes required for the effective technology transfer intended by current federal regulation. Federal and private funding agencies and industry should jointly undertake a suitable cost-benefit analysis and explore mechanisms to enhance the efficiency both of funding shared resources and of sharing resources. A major argument for the sharing of resources is the enhancement of both the effectiveness and the efficiency of doing research. To justify funding of resource sharing, it is necessary to be able to document the savings achieved. The capital investment needed and the demand for the product will help determine the number of and placement of facilities. For instance, synchrotrons by their very nature will be limited in number, and the same is likely to be true for primate centers. Culture collections may offer economy of scale, which would serve to limit their numbers. Because of the growth of economic nationalism and to avoid unnecessary duplication, the world scientific academies should convene to identify barriers to sharing resources across national boundaries and should develop mechanisms to overcome them. Ideally science is international. Historically, barriers to exchange of ideas, results, and reagents have resulted from concerns of national security. More recently, economic security has become a more prominent component of national security, and science has come to be appreciated as a major contributor to economic well-being. Appropriate user or sample fees and ground rules for partnerships between industrialized and developing countries demand attention. New culture collections being established in Germany and Japan will be totally funded by the government, raising concerns both about

OCR for page 71
--> unnecessary duplication and about the possibility of restrictions on the sharing of reagents in the future. Also of concern is the establishment of universal rules for the protection of intellectual property and a commitment to adhering to such rules. The overall issue demands rapid action on the part of the scientific community to forestall decisions at the national level that may be difficult to reverse. Because the private sector will continue to have a major impact on resource sharing, representatives from industry, nonprofit institutions, and funding agencies should be brought together to work toward solutions of current problems such as the following: • Overreaching claims on future ownership of inventions by providers of shared resources and research tools The major question is at what point the original provider no longer has a legitimate claim. This includes issues of how far reaching the licensing rights of the provider are and how long the sharing of a resource entitles participation as a full collaborator. • Competition between private-sector activities and public shared resources At what point should the distribution of a scientific resource be done by the private sector. Currently there is concern about "cherry-picking"—allowing public resources to do the hard work of development and quality control, only to have private businesses undercut these costs by taking advantage of the work done by the public resources. • How to protect the research exemption for licensed intellectual and tangible properties To what extent should there be a distinction between the use of resources for nonprofit research as opposed to work done for commercial development? If such a distinction should be made, how can that be achieved and what should it entail?

OCR for page 71
--> • Impediments to biomedical research and education caused by confidentiality requirements Have confidentiality requirements actually impeded research? Have they done damage to collegiality? To what extent is the lack of sharing caused by commercial concerns versus a more general unwillingness to share? A cost-benefit analysis should be conducted to evaluate the possible impediments to resource sharing caused by government regulations. The major considerations should be the extent to which such regulation actually contributes to the desired end, whether the desired end could be achieved in a more economical manner, and finally, whether the benefits really are commensurate with the costs.

OCR for page 71
This page in the original is blank.