Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7 Implementation and Management Strategies Earlier chapters have outlined the scientific strategies that seem most reasonable for genome mapping and sequencing studies and have argued in favor of an intensive effort to characterize the human genome in detail. In those chapters the committee has discussed the technological advances required for completion of the different human genome maps desired: a genetic map with REAP markers spaced at about 1-cM intervals, a physical map of expressed genes (cDNA map), an ordered DNA clone collection that covers the entire genome, and physical maps of increasing resolution culminating in the DNA sequence. However, such a large endeavor requires a degree of organization and coordination that has no precedent in the biological sciences. The committee-has argued specifically against establishing at this time one or a few very large production centers to carry out this project. Yet a concerted approach to project management will be required, and many of the benefits of a singular effort will be lost if results and materials are not quickly and efficiently shared. The committee's recommendations immediately raise a number of obvious questions, such as, How should such an effort be funded so that its quality and success are ensured? How can the efforts of the different laboratories best be coordinated? How can the scientific community guarantee ready accessibility to all the information and materials to be generated? This chapter deals chiefly with the com- mittee's attempts to answer these important questions. 86
IMPLEMENTATION AND MANAGEMENT STRATEGIES FUNDING A HUMAN GENOME PROJECT Projects with the Potential to Make Substantial Technological Improvements in Genome Analysis Should Receive Top Priority 87 Mapping and sequencing the human genome poses a severe tech- nological challenge to our present abilities. A complete restriction map has recently been produced for the genome of the bacterium Escherichia colt, which is only No the size of the human genome (Smith et al., 1987, Kohara et al., 19871. In addition, the human genome is 20,000 times as long as the longest continuous stretch of DNA sequence of 150,000 nucleotides thus far produced in a single laboratory project. Interpreting the large amounts of information produced by large-scale DNA sequencing will be an even more formidable task. Thus, the human genome project is more ambitious by orders of magnitude than any single biology project completed thus far. Technological advances are needed in many different areas, and determining the most efficient technology to obtain the desired knowledge requires further research. In view of the current situation, the committee recommends the establishment of a competitive grant program specifically focused on improving in 5- to 10-fold increments the scale or efficiency of mapping and sequencing the human genome. These grants would be designed to support work that is more technologically oriented than most ongoing biological research. For example, a project that aims at sequencing a fragment containing a contiguous segment of a million nucleotides or more might qualify for support, as might a project aimed at developing and testing an entirely new approach to DNA cloning or sequencing. In contrast, a project that aims to use standard approaches to isolate and characterize a single interesting human gene of 10,000 nucleotides would be more appropriately funded elsewhere. Both Small Research Laboratories and Larger Multidisciplinary Centers Should Be Encouraged Nearly all the major methodological breakthroughs that have driven the modern revolution in biology have come from the efforts of small research laboratories, frequently those led by young investigators in the early stages of their careers. This trend will no doubt continue, and small groups that are already active in developing or improving relevant technologies should be supported by this project, irrespective of the particular genome that they are characterizing. There is also much to be said for the establishment of multidisci
88 MAPPING AND SEQUENCING THE HUMAN GENOME plinary centers, in which 3 to 10 research groups in a single facility, each with a different but related focus, share equipment and personnel and collaborate to accomplish a larger goal than any single group could readily achieve. For the most part, these should be located at universities or research institutes that can help provide the projects with a constant influx of new people and ideas. Efficient mapping requires the coordination of a large number of different experimental and computer techniques. It also demands flexibility in developing and incorporating new technologies. The methods will probably evolve very rapidly, and the strongest and most efficient mapping efforts will provide their first critical tests. For these reasons, a substantial portion of the human genome mapping effort should probably be organized into medium-sized research centers, each with ongoing activities in both development of techniques and actual mapping and each with a reasonable fraction of the various technologies in place or under development. Such centers can set the stage for subsequent large- scale DNA sequencing projects. The sequencing effort will benefit from close contact with mapping efforts, some of which provide the samples to sequence, while others provide the framework to organize the sequence data generated and to assist in its interpretation. The Establishment of a Single Large Production' Center Is Not Advisable at Present The committee believes that it would not be wise to confine an activity such as mapping or sequencing to a single, large center at present. The task needs to be organized and coordinated, but it does not need to take place in a single location or laboratory. Unlike many physics projects of this magnitude, this large-scale biology project can be subdivided along several different lines. For example, an individual chromosome could be mapped by one laboratory, but even this is not necessary since available methods permit one to work efficiently with just a section of a chromosome. Similarly, while restriction, genetic, and DNA-clone-collection mapping must be con- ducted in parallel, they do not necessarily need to be done by the same investigators, inasmuch as materials generated by one method can readily be transferred for use in other laboratories. There are strong technical and intellectual advantages to dispersing much of the mapping efforts among medium-sized, multidisciplinary centers (each with perhaps 30 to 100 persons). If adequately funded, these units would be large enough to accommodate biologists, chem- ists, physicists, and engineers with diverse skills and backgrounds, thereby achieving the critical mass necessary for an effective approach.
IMPLEMENTATION AND MANAGEMENT STRATEGIES 89 At the same time, available resources should make it possible to establish several centers with the same goals, each competing with other centers to advance the technology. Such competition is healthy and productive, and it permits subsequent judgments to be made on relative quality, allowing additional resources to be directed to the most successful units. Finally, the dispersal of these technology centers into existing universities and research institutes will allow members of the center to work closely with a large number of other scientists. This interface should be of great value to both groups, and it will ensure that the human genome effort will have the strong support it needs from the scientific community at large. Decisions for Funding Should Be Made by Peer Review The committee envisions that funding would initially be offered in the form of grants awarded for 3- to 7-year periods. It is imperative that these grants be awarded solely on the basis of scientific merit, as judged by panels of peer reviewers selected for their judgment and scientific expertise. The committee specifically recommends the form of review that has been routine for many years at the National Institutes of Health (NTH), in which the reviewers meet to discuss and debate the merits and faults of each grant application. Each application is then ranked relative to the others by the assignment of a priority score, which is forwarded to the individual NIH institute that distributes the research funds. Generally, grants are awarded in the order of their priority scores, until the allotment of funds for that cycle is exhausted. As technologies mature, production units, such as contract orga- nizations or dedicated centers will be required in the human genome project. For example, RAP mapping is already a relatively mature technology. An REAP map at higher resolution can be attained mainly by applying current methods on a larger scale. Such endeavors are appropriately supported by contracts rather than by grants. The central facilities that collect and distribute information and materials should also be supported by contract. Such efforts should also be subject to continuing peer review, both for technical competence and to ensure continuing coordination with the overall effort to map and sequence the human genome. If a human genome project is funded by several separate U.S. government agencies as well as by private funds, an effective reviewing body will be needed to avoid excessive duplication of effort and to oversee cooperation between research groups. The committee rec- ommends that the same body also ensure a uniform standard of peer review.
90 MAPPING AND SEQUENCING THE HUMAN GENOME The Human Genome Project Requires New and Distinctive Funding of About $200 Million per Year To create the multidisciplinary centers suggested, new laboratories will need to be built, equipped, and staffed. Universities cannot be expected to provide the necessary resources without a major new source of funding from outside. The estimated cost of an effective project to map and sequence the human genome is $200 million per year. (This level of funding would be reached only during the third year of the project, with the first 2 years having lower levels of funding to allow a scale-up to an effective project.) The money might be spent roughly as follows. In the first 5 years of the project there might be about 10 medium-sized, multidisciplinary Protons ~nr1 m~nv smaller research ground working, with perhaps half of the projected total of 1,200 individuals in the multidisciplinary groups. Each professional researcher costs about $100,000 annually in pay and support (the standard number used in the biotechnology industry); thus this cost would be an estimated $120 million annually. =~~~r~ -----a ----I---- ----I---- =---¢~ --on Construction and equipment might cost about $55 million per year, the stock center, the data management center, the quality control effort, and the Scientific Advisory Board (see below) might cost approximately an additional $25 million per year. As the project proceeds, annual construction costs will decrease, but the number of indivicluals participating in the effort may increase to about 1,500. The committee's possible scenario for the project divides the effort into three 5-year periods (l, It, and IlI). During each period, mapping and sequencing efforts five times as complex as the next lower numerical designation would be undertaken at constant cost, reflecting five-fold increments in technological sophistication. Several points should be stressed. · Attaining two successive fivefold increases in the technology for mapping and sequencing is an ambitious undertaking; if it is to succeed, a major effort must be expended in developing the required technology. · The cost for sequencing must include preparation of the DNA samples. New methods of DNA subcloning and processing will have to be developed (or present ones automated) to attain the desired costs. · DNA clones from an ordered DNA clone collection will be sequenced, thereby producing large, contiguous stretches of DNA sequence that are immediately useful; isolating the last 10 to 15 percent of these clones to fill in gaps in the map may be as expensive as
IMPLEMENTATION AND MANAGEMENT STRATEGIES 91 isolating the clones that cover the initial 85 to 90 percent of the genome. · An ambitious effort of this type will require the recruitment of scientists with extensive experience in mapping and sequencing. The multidisciplinary centers supported by the project will presumably play a key role in training new independent scientists who can participate. The committee believes that the training of young scientists in the development of technology as well as its applications would be of the major benefit to the biological community. A major objective of the human genome project would be to achieve an annual sequencing capacity of ~ billion nucleotides through the combined efforts of a modest number of centers by the year 2000. Once this ambitious goal is reached, it would be realistic to complete the entire human genome sequence, and powerful comparative studies on human polymorphisms and evolution would become possible. Funding of the human genome project must not be at the expense of currently funded biological research. The essential purpose of the human genome project is to provide a resource to be used by biomedical scientists to accelerate the understanding of human biology and the application of this knowledge to human health. It would therefore defeat the purpose of the project if biomedical research and training were weakened by diverting funds from individual research programs or training grants. Major advances being made in other aspects of biology are expected to form the scaffolding required for interpreting and utilizing information resulting from this project. Mapping Efforts Should Be Accelerated and Coordinated Most biologists feel that mapping the human genome is a valuable and attainable objective. At the present rate this goal will not be reached for many years. The committee strongly believes that this effort should be accelerated. Funds should be invested in projects that increase our technological skills, but large-scale mapping should begin even with present-day techniques. It should be possible to complete an RAP map within 5 years with the investment of more money and the encouragement of coordinated efforts. Several types of physical maps can also be completed in a similar length of time. The Sequencing Effort Should Evolve andt Grow with Time Most of the human genome will probably become available in the ordered DNA clone collection as a result of the combined efforts of several multidisciplinary centers. If the envisioned support is forth
92 MAPPING AND SEQUENCING THE HUMAN GENOME coming, the ordered DNA clone collection could be completed within 10 years. In principle, sequencing major blocks of the human genome could begin as soon as any contiguous area of a chromosome has been accumulated in the ordered DNA clone collection. Decisions for a major push on bulk sequence data collection, as distinct from the envisioned pilot projects that push technology development, would depend on how fast the new sequencing technologies develop. Several relatively small genomes should be sequenced early in the project. Such genomes include that of yeast (0.5 percent of the size of the human genome), the nematode Caenorhabditis elegans (2 percent of the human), and the fruit fly Drosophila melanogaster (3 percent of the human). It is important to give these projects high priority, since they deal with widely studied experimental organisms, which, along with the mouse, provide the most important of the many mode! systems that will be required for interpreting all the sequence data that will be collected on humans. International Collaboration on the Project Is Desirable There is a strong tradition of international cooperation in the biological sciences that has greatly speeded the rate of scientific progress in the past. This tradition must continue in any human genome project. Some portion of research on the human genome is bound to be done outside the United States, mainly in Europe and Japan. One project already started in Japan is supported by major industrial companies, which intend to automate DNA sequencing at the rate of 1 million nucleotides per day. In Europe, developments in semiautomatic DNA sequencing at the European Molecular Biology Laboratory in Heidelberg have resembled those at the California Institute of Technology. Work on new methods for physically mapping complex genomes, including the human genome, is progressing in Cambridge and London. In addition, extensive research on human genetics is under way in several European countries. The reference set of 40 families collected by the Centre d' etude du Polymorphisme Humain and used throughout the world for mapping RF~Ps was established in Paris, from which reference cells and DNA are distrib- uted worldwide. GenBank in the United States currently shares on a 50:50 basis the collection and entry of DNA sequence data with the EMBE Data Bank in Heidelberg. These examples suggest that the United States does not and cannot expect to monopolize information and innovation in this field. More- over, the initiation of a human genome project in the United States will probably not deter work in other countries, but rather will
IMPLEMENTATION AND MANAGEMENT STRATEGIES 93 stimulate it. Given this assumption, the importance of past traditions, and the magnitude of the task of mapping and sequencing the entire human genome, every effort should be made to enhance the existing contacts between U.S. laboratories and those overseas, so as to speed the work. Indeed, we believe it will become necessary to have some major organized mechanism for international cooperation. In partic- ular, its objective would be to collate data and ensure rapid accessibility to it, as well as to distribute materials, such as cloned DNA fragments. MANAGING A HUMAN GENOME PROJECT The human genome project presents complexities of organization and management at both the scientific and policy levels. At the biological level, the project differs from conventional biological re- search in that it must be coordinated in terms of mapping, definition of overlapping cosmids, distribution of DNA clones, sequencing, technology development, and data base design. At the policy and funding level, a number of governmental and private foundations will probably be involved. For these reasons, it is imperative to design a management system that will provide oversight, coordination, review of progress, and forward planning. The committee was convinced of the need for strong leadership for the project, and presents three possible management plans. We also recognize that the management of the human genome project may need to evolve as the project evolves, as have management mechanisms for similar projects with federal research support. Three Possible Organizational Plans Briefly summarized, each of the possible organizational plans (designated A, B. and C) includes a Scientific Advisory Board, but differs in the administrative and funding leadership. In plan A a single federal agency serves as the lead for the project. This agency, which would be assisted by a Scientific Advisory Board composed of experts who provide scientific advice for the project, would be responsible for all aspects of the project. In plan B an Interagency Committee, consisting of representatives of the National Institutes of Health (NTH), the Department of Energy (DOE), the National Science Foundation (NSF), and other federal agencies interested in the project, would be responsible for all aspects of the project. The Interagency Committee would' be assisted by a Scientific Advisory Board that would provide advice on the project, as in plan A. In plan C, an Interagency Committee would have the ultimate responsibility for the coordination and funding of the activities to be
94 MAPPING AND SEQUENCING THE HUMAN GENOME supported, while a single agency would be responsible for the daily administration of the project. In this plan, a Scientific Advisory Board would provide advice to both the Interagency Committee and the administrative agency. Each of these management plans has advantages and disadvantages. Although the primary charge to the committee was not to develop an organizational plan, and decisions on management organization include considerations outside the committee's areas of expertise, a majority of the committee members favors plan A, which therefore is presented in the greatest detail. Organizational Plan A: A Lead Agency and a Scientific Advisory Board The Human Genome Project Should Be Assigned to a Major Federal Agency In this plan the human genome project would be sited within a federal agency as an independently funded endeavor. This would place both the responsibility and the operation of the project within a single unit. Already there are aspects of the human genome project being supported by three federal agencies, N]:H, DOE, and NSF. Each of these agencies has the potential capability to provide an effective home for the project. Although the committee did not fee! it was its role to designate a lead agency, it did discuss some of the merits of each agency. The NTH is the major agency today supporting research on the structure and functioning of DNA. It is also the agency with the mandate to foster biomedical research in the United States. It has a long history of successful support of peer-reviewed extramural research, successful intramural laboratories on the NTH campus, and has been involved in the oversight of large projects in biology, such as the viral cancer program. Recently, the NTH has shown interest in establishing a human genome project. DOE has successfully managed many impor- tant projects for the physics community, has supported some life science programs, has extensive computer facilities for data manage- ment, and has expressed strong interest in overseeing the human genome project. The NSF has been involved in the development of technology and instrumentation relevant to the human genome project, in the general support of basic biological research, and has a well- established peer review system. Locating the project within one agency does not mean that all the funding for the project would flow from it or that scientists associated with another agency would not be able to obtain funding. For example, even if DOE were not the lead agency, scientists at the national
IMPLEMENTATION AND MANAGEMENT STRATEGIES 95 laboratories and DOE grantees should be eligible to compete for research support relevant to the human genome project. The same would apply for scientists at the NTH and NTH grantees, if NTH were not the lead agency. Scientific Advisory Board Although the lead agency would have the ultimate authority and responsibility for the funding and admin- istration of the project, our committee believes that the overall scientific oversight of the project should draw upon the experience and be guided by a Scientific Advisory Board (SAB) made up predominately of scientists with expertise in the methods and goals of the project, and chaired by a full-time chairman who is a distin- guished scientist. We recommend that the members of the SAB be appointed by the lead agency for staggered terms. The chairman should be appointed for a fixed term with a possibility for reappoint- ment and should be provided with a full-time staff sufficient to carry out the directives of the lead agency and the SAB. The role anticipated for the Board is somewhat stronger than that of a typical scientific advisory board. The major responsibilities of the SAB include: · To facilitate coordination of the efforts of the many laboratories that are expected to participate in this effort. · To help assure the accessibility of all information and materials generated in the project by advising on the oversight of the data center and the stock center and recommending contracts where appropriate. It would oversee formation of standard terminologies and reporting formats so that the large body of information to be obtained can be readily communicated and analyzed by the entire scientific community. · To monitor the quality of research by helping to assure a uniform standard of peer review. · To suggest mechanisms for strict quality controls on the sequence and mapping data collected. · To promote international cooperation, serving as a liaison to projects outside the United States regardless of their funding sources. · To make recommendations concerning the establishment of large sequencing endeavors, thereby balancing focus with breadth. · To publish periodic reports stating progress, problems, and recommendations for research. The Scientific Advisory Board Should Provide Advice on the Peer Review Process and on Coordination of the Project A human genome project cannot succeed unless the various mapping and sequencing
96 MAPPING AND SEQUENCING THE HUMAN GENOME efforts are coordinated. Standards must be set and the rapid distribution of materials and information must be facilitated. The failure of a unit to coordinate its efforts with others should jeopardize the support of the unit not in compliance. The lead agency and SAB should work closely in developing and implementing a high standard of peer review. After the committee considered several options, it concluded that this system would function best if the SAB were involved in monitoring the evaluation of proposals for funds. After consideration of the advice of the SAB, the lead agency would then fund selected applications. The Scientific Advisory Board Would Require Funding To be effective, the SAB must be adequately funded. Although the lead agency would provide much of the money for the SAB, private foundations and institutes should be encouraged to help support it. An appropriate mechanism for merging private and federal funds for this purpose would have to be developed. If, in addition to its role as scientific advisor and coordinator, the SAB is assigned such tasks as oversight of peer review panels, funds will have to be provided to it for these purposes. Organizational Plan B: An Interagency Committee and Scientific Advisory Board The Interagency Committee Three government agencies can po- tentially play leading roles in the human genome project: The National Institutes of Health (NTH) because of their central responsibility for human biomedical research and their exemplary peer-reviewed extra- mural grant programs; the Department of Energy (DOE) because of its interest in the project and experience in data management and the management of large-scale projects; and the National Science Foun- dation (NSF) because of its commitment to technology development and basic research support of biology across the discipline. Private foundations and institutes may also be interested in involvement in the project. In this plan the above three governmental agencies would form an Interagency Committee (IAC) that also includes members from other agencies interested in the project. The chairmanship of the IAC might rotate among the three principal agencies (NIH, DOE, and NSF) involved in the project. The IAC would be responsible for overall administration of the project, including funding of research programs and supporting serv- ices; administration of a common peer review process, the stock
IMPLEMENTATION AND MANAGEMENT STRATEGIES 97 center, and the data center; and the appointment of and response to a Scientific Advisory Board (SAB). Because funding for this project may come from several agencies it is important that the Interagency Committee be responsible for the coordination of the funding. Each year the committee should develop a total project budget and determine what the contributions of each of the agencies would be to the project. Representatives of each agency can then request the funds required for the project from the administration and the respective appropria- tions committees. The Scientific Advisory Board would assume a role similar to that outlined in organizational plan A, except that it would of course advise the Interagency Committee rather than a lead agency. Organizational Plan C: Interagency Committee, Ad~mir~istrat~ve Agency, and, Scientific Advisory Board In this final proposed plan aspects of organizational plans A and B are combined to form a three-part administrative structure. The Interagency Committee As discussed under organizational plan B. an Interagency Committee would be established to oversee the project. It would be ultimately responsible for the coordination and funding of the activities to be supported, the administration of the peer review process, the research program, the stock center, and the data center. It is expected that this committee, which would provide the administrative and the funding lead for the project, would pay close attention to the recommendations of the Scientific Advisory Board and the administrative agency. The Administrative Agency In this plan, one of the member agencies of the Interagency Committee would act as the administrative agency responsible for the daily administration of the project. This agency would be involved in the administration of funds and other administrative aspects of the project, such as the operation of the stock center and the data center. Questions and inquiries about the project would be directed to this agency which would serve the important role of clearinghouse for the effort. This agency would help to guarantee that the project is well-run, and that the necessary details in the operation of this large-scale project are completed. The administrative agency would work closely with the SAB in developing and implementing a high standard of peer review. This agency would arrange the administrative details of the peer review process, while the SAB would monitor the initial evaluation of grant and contract proposals. After the assignment of a priority number to each grant and contract, the Interagency Committee would select the
98 MAPPING AND SEQUENCING THE HUMAN GENOME applications to be funded. The Scientific Advisory Board in this organizational plan would have similar responsibilities to those outlined in plan A, but would advise both the Interagency Committee and the administrative agency. REFERENCE Kohara, Y., K. Akiyama, and K. Isono. 1987. The physical map of the whole Escherichia cold chromosome. Cell 50:495-508. Smith, C. L., J. G. Econome. A. Schutt, S. Klco, and C. R. Cantor. 1987. A physical map of the Escherichia cold K12 genome. Science 236:1448-1453.