1. Sequencing Additional 3' ESTs to Facilitate Mapping and Gene Identification

A. New Library Generation

    ¨ Use Normalization/Subtraction

    ¨ Cost: $0.5 Million Duration: 2 Years

A. EST Sequencing

    ¨ 200,000 3' EST Sequences

    ¨ For Unique Clones - 5' EST Sequences

    ¨ Cost: $1.5 Million Duration: 3 Years

1. Full-Length cDNA Sequencing - Major Priority

A. Technology Development and Generation of Representative Full-Length cDNA Libraries

    ¨ Cost: $5 Million Duration: 3 Years

A. cDNA Sequencing

    ¨ 3% of Genome (estimated coding sequences)

    ¨ 90 Mb

    ¨ Cost: $45 Million @ $0.50/bp Duration: 3-5 Years

1. Prioritization of Strains/Tissues for Library Generation and Quality Control

    ¨ Community Steering Committee

Breakout Group C: Genetic Mapping

The breakout group concluded that the highest priority for the future of genetic mapping in the mouse was a SNP map. This would greatly simplify genotyping for several different purposes, e.g. positional cloning, identifying QTLs, analysis of LOH. It was also considered to be of very high priority for the SNPs to be in the public domain as quickly as possible, so that the academic community could have access to them. The group distinguished two aspects of the SNP question, SNP discovery and SNP deployment (scoring).

A specific recommendation was made with respect to SNP discovery. Initially, a relatively small number of SNPs, on the order of 1500-2000, should be generated. This would make it likely that in any particular pair of strains used in a cross, at least 300 - 400 SNPs would be informative, allowing a genome scan at about 5 cM resolution. In the first instance, SNPs should be chosen to give a uniform distribution across the genome, and to be useful for about 10 different diverse strains. It was estimated that a project of this magnitude would cost about $1-2M and it was recommended that two such projects be supported, in order to provide healthy competition for the SNP developers. It was thought that this could be done in two years, at which time the value of the SNP approach could be determined and, if warranted, a more dense collection could be developed. . It was considered important for the initial set of SNPs to be integrated with existing maps, but to leave the mechanism for doing to be determined by competition

With respect to deployment, it was felt that the development of the initial set of SNPs did not imply any specific technology for further use. Rather, the group recommended the support, chosen through a competition, of a variety of new approaches to scoring, that would lead to widespread use by a large number of small laboratories, as well as by the major users. It was also noted that technologies for scoring SNPs would also be applicable to the detection of induced point mutations.

As for the choice of strains, the group thought that it would be most appropriate for recommendations about specific strains to be left to a small committee of experts, that would include representatives from the NIH, from the Jackson Lab, and from Europe. It was felt that a list developed by this group should then be circulated to the wider community for a period comment and validation. Several factors were suggested for consideration in selecting the strains: genetic diversity, usefulness in studying a variety of phenotypes, usefulness in mutagenesis and knockout screens.

A second priority for genetic mapping resource was the development of additional sets of consomic strains. Currently, one such set is being developed, a set of A/J chromosomes onto a B6 background. The group thought that the recombinant consomic technology had the potential for major advances in QTL mapping, and that its usefulness would extend well beyond the traditional mouse genetics community in the sense that it would allow physiologists, developmental biologists, neurobiologists, etc. to rapidly and simply map genes that determine or modify phenotypes of interest. In this case, the recommendation was that initially NIH support the development of about 3-5 sets of particularly useful strains. Although the ultimate choice of strains should be made in a competition, it was felt a matched set of 129 and B6 consomic strains would be extremely useful. It was estimated that each set would cost about $500,000.

A third recommendation was to characterize more inbred strains by typing them with a large number of SSLP markers. It was thought that existing inbred strains provide a rich resource for the analysis of complex traits, but many are underutilized because they are not well genetically characterized. It was estimated that a one-time project typing about 6000 markers on 50 strains would cost about $1M and take about a year, and had the potential for opening up many more strains for the study of mutations, QTLs and modifiers.

Finally, a radiation hybrid panel that would allow markers to be ordered at higher resolution than the existing RH panel would be of value. Again, generating this resource would be a one-time effort, and was estimated to cost about $500,000.

BREAKOUT GROUP D: Genomic Sequencing

(Revised)

How will the mouse sequence be used?

This must be understood to address the questions below. Several different areas were identified:

1. positional cloning -- 1-2 BACs for 25-50 projects a year

2. region specific mutagenesis - 3-5 Mb

3. comparative sequences to define other elements, breakpoints etc.

4. reverse genetics -- small targeted regions

5. reference tool to interpret human sequence

For small projects directed at getting specific genes, the group concluded that means should be devised to enable small groups to obtain the needed sequence. This might, for example, involve skimming, followed by more accurate sequence for the gene of interest. This scale of sequencing should be brought within reach of most labs. The other projects however require more systematic sequencing on a larger scale and would provide a long term resource to the community. Comparative analysis of mouse and human sequence was the most compelling use in the eyes of the group.

What quality for the sequence?

This questions pits coverage versus completeness. There was considerable discussion of alternatives, including options such as 2x shotgun of BACs across the genome, or leaving a limited number of gaps to reduce costs. However, there was greater enthusiasm for a commitment to significant amounts of mouse sequence in the next few years and to get the whole mouse genome done in a reasonable time (vide infra). The group concluded that the temporary advantages of quicker, broader coverage were outweighed by the added costs incurred in achieving the ultimate goal of the complete reference mouse sequence. Accordingly, there was broad agreement that the mouse sequence standards should be the same as for human.

What is priority of completing the mouse relative to the human?

The greatest value would result from parallel efforts on human and mouse. However, practical limits of resources and capacity and the commitment to complete the human genome by 2005 will necessitate that priority be given to human sequencing. Nonetheless, the greatly added value that the mouse sequence will give to the human make it imperative that mouse sequencing not lag far behind human. It is important to begin a significant effort on mouse sequencing now, and expand that as rapidly as resources, capacity and the commitment to sequence the human allow. For example the group felt that it was more important to do more mouse sequence than finish the human early.

What priorities should be followed in the choice of mouse genomic regions to be sequenced?

If individual labs are enabled to do sufficient sequencing to support positional cloning efforts, they will establish their own priorities. For larger systematic efforts, it would make most sense to focus on regions of the mouse genome syntenic to those already sequenced in human. Larger regions of particular biological interest should also be considered. Selection of specific regions that match these broad guidelines is best left to individual labs in dialog with larger scale sequencing groups. A tool similar to the Human Genome Sequencing Index maintained by the NCBI should be constructed for the mouse genome and the syntenic regions of the human and mouse map should be linked. Such a tool will be useful to mouse researchers in determining what groups are mapping and sequencing the human genome in regions syntenic to their regions of interest and it will help to track and coordinate the mapping and sequencing of the mouse.

Is strain a consideration for generating a reference mapsequence?

Overall, the group felt there was little scientific reason for favoring one strain over the other. The one mandate was that the sequence be derived from a single well characterized strain, and not be a mosaic of different strains. The one practical consideration is that the BAC clones from 129 are already heavily used. Whatever strain is chosen, the selection should be made known, and the clones and other resources must be readily available.

What should be the priorities for resource generation in the sequencing area?

Highest priority should be given to the development of a sequence ready map. The estimated cost of building a BAC overlap clone map was estimated to be approximately $7M. Initial fingerprinting, contig building and chromosomal positioning could be done over three years. Efforts at closure would extend over two more years.

The next priority is to begin systematic sequencing of the mouse genome. Every effort should be made to increase available sequencing capacity to allow the amount of mouse sequence completed annually to grow from the 10-12 Mb in the next year to 400 Mb in 2005. The goal should be to obtain 1 Gb of mouse genomic sequence in this period. The remainder of the mouse genome should then be completed in the next three years. Thus by 2008 a complete reference mouse sequence should be available.

Finally means should be found to facilitate the sequencing of small regions (100-200 kb) by R01 funded labs. Equipment seems less of a problem than access to methods and informatics support to handle the relatively large data sets. Perhaps a resource similar to that in the UK could be established.

Breakout Group E: Targeted Mutagenesis

Targeted mutagenesis is a well-established technology and widespread in the community. Its value for analyzing gene function is undisputed. Standard approaches for generating null mutations are now being complemented by generation of allelic series of mutations and tissue-specific and induced mutations, which further enhance the utility of this approach. The group focussed on issues of standardization and resource availability across the community, enhancement of emerging technologies and the importance of new tools for in depth phenotypic analysis to fully exploit the mutations being generated.

A. Standardization of resources for gene targeting.

There was a strong consensus for use of a standard genetic background for gene targeting. Current use suggests that 129SvEvtac is the line of choice, for stability of lines and ease of maintaining inbred lines. Validated lines exist on this background.

- Support a mechanism for affordable, reliable distribution of validated 129SvEvtac (Hprt-)ES lines to the community.

Cost: $100k/yr, ongoing

It was also recognized that there is a need for validated ES lines from different strains for specialized uses. A better understanding of the biology of ES cells could assist in such derivation.

- Support establishment of validated ES cell lines from other strains and stable female cell lines.

Cost: $500k/yr for 3 years

There is a need to study mutations on different strain backgrounds, requiring backcrossing from 129 to other strains. Speed congenics assist in this approach, but are not easy for the average lab.

- Develop set of SSLPs for easy generation of speed congenics onto standard backgrounds (some may be available already)

Cost: $100k/yr for 2 years.

Issues of targeting technology were considered. The rate-limiting step in targeting is making the vector. The availability of BAC libraries in the 129SvEvtac strain is very important in this regard. One study suggests that it may be possible to use the BACs directly to make targeting vectors, which would then have very long regions of homology to enhance targeting efficiency.

- Test the feasibility of direct generation of targeting constructs in BACS and the possibility of targeting by zygote-injection.

- Cost: $300k/yr for 3yrs.

Possible strategies for generating genome- wide directed mutations were considered, such as sequence-tagged gene trap libraries or libraries of ENU-mutagenized sperm. The group did not support a new public-domain genome-wide gene trap effort and felt that the market place would determine the usefulness of Lexicon's current library. Any large-scale ENU screening center should be encouraged to archive sperm and DNA for possible sequence-based screening.

B. Conditional targeting strategies

The power of tissue-specific and inducible mutagenesis for dissection of the full roles of a gene is enormous. There was a strong feeling that the Cre-Lox system was working and that the main impediments to further use are the availability of fully characterized lines expressing Cre in a defined spatial and temporal manner. It was felt that many of the problems of mosaic Cre expression could be overcome by insertion of Cre into the genome so that it is controlled by endogenous regulatory elements. In addition, good ubiquitous lox-stop-lox reporter lines to validate Cre expression are needed. For inducible targeting, the nuclear receptor- Cre fusions show considerable promise and are being worked on extensively.

- Support effort to make versatile ubiquitous excision reporter lines and lines of mice expressing Cre in defined spatial and temporal patterns by knock-ins or enhancer trap approaches

Cost: $2M/Yr for 4 years

There is a clear case for a centralized facility to distribute the Cre lines and to create a database of the Cre lines. This should be costed under the resource facility for mutant distribution. There is still an outstanding patent issue with Dupont, which could inhibit distribution of this resource. The group considered the site-specific recombinase technology to be so important that, if this issue is not resolved soon, there must be an immediate switch to an alternate enzyme system to avoid wasting effort.

C. Phenotyping

This is a major challenge in all mutagenesis efforts. There is an urgent need for improved technologies for mouse physiology, pathology, antibody markers, reporter mice, expression arrays, behavior analysis, etc. It was felt that individual Institutes will need to invest in this to obtain the full value of the mutant resource for understanding mouse biology as it impacts on human disease.

- That the NIH convene a meeting this year to discuss enabling technologies for mouse phenotyping in different organ and tissue systems

D. Additional issues

1. Need for continued support for facilities to store and distribute mouse mutants.

2. Importance of cryopreservation for mutant stocks.

3. Need for mutation database, fully annotated with phenotypic information.

4. Need to address animal husbandry and whether current costly procedures are necessary for mouse welfare.

Breakout Group F: ENU Mutagenesis

Rationale: Phenotype-driven mutagenesis screens provide an important approach to understand the function of genes. As the genome project progresses and the sequence of more human and mouse genes are determined, the function of a large number of genes will not be predictable by sequence and expression alone. To understand the function of genes, mutations have provided a powerful approach to understand to this problem. In the mouse, the supermutagen, ENU, produces forward mutation rates (~1/650 per locus per gamete) sufficient to perform genome-wide mutagenesis screens at significant coverage so that, on average, mutations can be obtained in almost any gene. Phenotype-driven ENU mutagenesis screens provide a complimentary approach to gene targeting methods because both gain-of-function and loss-of-functions alleles can be obtained and importantly no pre-existing knowledge of the target genes is required. This forward genetic approach has been widely used and proven in model genetic organisms, however, in the mouse large-scale screens need to be performed in order to determine whether ENU mutagenesis can be applied broadly to understand gene function in the mouse. There was clear consensus from experience within the Breakout group and others that ENU screens are highly efficient and effective in isolating mutants for particular phenotypes. In order to apply this approach in a more systematic manner, the group recommends that a number of new initiatives should be launched (which are described below).

Ongoing large-scale ENU mutagenesis screens

Currently there are two large-scale ENU screens that have been initiated in Europe. Rudi Balling at the GSF is conducting an integrated screen of 40-50,000 mice for a number of dysmorphic and blood-assayable biochemical phenotypes. Steve Brown at Harwell is conducting a similar size screen of neurobehavioral phenotypes. In the first 6-12 months of the screens, significant numbers of mutants have already been obtained (about 150 mutants from 14,000 mice in a one generation dominant screen at GSF, and about 60 mutants out of 3000 mice screened at Harwell). Thus, the feasibility of large-scale "Center-based" integrated ENU screens appear feasible, efficient and productive.

Questions to the Breakout Group.

1. How can the efficiency of mutagenesis be monitered and improved?

In discussing the various screens, the group concluded that there is a significant variation in the mutations rates observed, strains used and breeding protocols employed. Experience at Oak Ridge (Monica Justice) and Madison, WI (Bill Dove and Alexandra Shedlovsky) suggests that strict quantitation of ENU doses is essential for achieving high mutagenesis rates (rates of 1/200 per locus per gamete). There is a need for protocol validation with different strains of mice (used for different phenotypes) and a need for dissemination and training of ENU mutagenesis procedures to make this approach more widely useful to the scientific community.

The group felt that adequate methods for assessing mutagenesis rates at the molecular level are not cost effective at this time. It would required scanning Megabase pair regions thoughout the genome to sample ENU-induced mutations adequately (the estimated molecular rate for ENU would be about 1 per 100,000 bp).

2. How can the cloning of point mutations fe facilitated?

The group strongly endorsed the recommendation of the structural genomics groups that a complete physical map of the mouse genome in BACs be made and that a high priority should be placed on full length cDNA cloning and sequencing (to high quality sequence standards).

3. What approaches are needed to improve the ability to identify recessive loss-of-function alleles?

The group endorsed ENU screens using targeted deletions in regions of the mouse genome that will have high priority for sequencing because of synteny with human, gene rich regions or regions of particular interest. In addition, large-scale recessive screens would be encouraged using the Centers of Excellence recommended below.

4. How can screening for phenotypes and validation of mutations be improved?

See recommendations below.

5. What the the priorities for chemical mutagenesis and what is the cost vs. the benefit?

Recommendations

I. The NIH should support systematic ENU screens in mice. Both Centers and individual investigators should be supported. A significant investment is required to test the feasibility of ENU screens as an approach to study the function of genes and pathways for broadly based phenotypes.

A. Establish Centers of Excellence for ENU screening that have a biological focus (with self-selected groups of investigators and consortia). These Centers would have:

¨ Mutagenesis core

¨ Phenotypic screening core (and technology development)

¨ Database core for dissemination of information

¨ Mapping core

¨ Sperm cryopreservation and DNA bank of G1 mice

¨ Mutant/sperm distribution provision

Establish at least 3 Centers of Excellence at $1-2M per year for 5 years.

A. Initiate an RFA to support individual investigators for other ENU screens that are not optimal for the Center format.

¨ Specialized phenotypes

¨ These R01's would have access to core facilities at Centers.

II. Initiate RFA to establish baseline characterization of a set of standard (consensus) phenotypes (behavior, neurological, motor, endocrine, biochemical, etc.) on a set of inbred strains of mice. This would provide a foundation of information for physiological parameters for strains of mice. The phenotypic characterization would be coordinated with the Centers and information would be disseminated by the database core facilities. In addition, the RFA should include development of technology for phenotypic screens in mice. For example, to transfer assays currently using the rat to the mouse. $1-3M per year for 5 years.

A systematic validation of ENU protocols on a standard set of mouse strains should be performed. ENU protocols and information should be disseminated.

$0.5M for 1 year.

BREAKOUT GROUP G:
Chromosomal Strategies, Insertions and Transgenes

1. What is needed to further the development of insertional mutagenesis?

An ideal insertional mutation system in the mouse should provide sequence-tagged Insertion of the germline on a genome wide scale. Such a system would facilitate functional genomics by providing a mutation at every locus, whose phenotype could be evaluated in the whole animal. Although efforts have been made to generate a P-element type transposon based in vivo insertional mutagen for the mouse, no such system is presently available. It is therefore recommended that high priority be assigned to obtaining an ES cell based resource that would provide a sequence-tag for each locus in the mouse genome. Such a resource would make it possible to scan the sequence-tag database for any gene of interest and to order the corresponding targeted ES cell line. This resource would permit the larger community of investigators to utilize genomic resources efficiently, and would be much more cost effective than the current effort to generate targeted knock-outs in individual laboratories, at an estimated cost of approximately $50,000 per knockout. Of course, this resource would only provide one null allele per locus, and would not obviate the need for ongoing targeted mutation of loci that are studied in depth. It would, however, be of enormous value to a large community of researchers. Making such a system available is of highest priority, and would be worth an investment of 10 M per year for several years. (Nearly this much is probably going into generating a few hundred knockout mice per year in small labs.)

2. What are the prospects for developing a collection of segmental deletions across the mouse genome?

Two ES cell based resources capable of generating deletions at marked sites throughout the genome are currently under development. The Merck Genome Initiative is supporting work in John Schimenti's lab to generate 500 ES cell lines with randomly inserted selectable TK inserts. The positions of the inserts will be determined by plasmid rescue of flanking DNA that can be mapped on a backcross panel. These ES cell lines will be described on a database and made freely available on request. Deletions from fixed insertion sites are also being generated by Alan Bradley at Baylor. Experience with these resources during the coming years will indicate whether additional resources will be needed in the future.

3. What is the value of analyzing polymorphic modifiers and QTLs, and how can their study be advanced?

Analysis of loci contributing to complex traits in the mouse was seen as an important future activity with broad biological applications. Many of the physical mapping resources discussed here will be important for QTL analysis. These include the generation of BAC libraries and maps, the development of SNPs, the generation of consomic lines of mice, the targeted ES cell library, and the development of expression arrays. All of these initiatives were strongly supported for their applications to QTL analysis.

4. What improvements in phenotyping are needed to facilitate functional genomics in the mouse?

A great deal of basic biological characterization of standard strains of mice will be required as background for assessment of phenotypic alterations in mutant mice. There is a serious shortage of expertise and manpower for sophisticated evaluation of mutant phenotypes. The following initiatives are proposed to address this problem:

a) A training program to support veterinary fellows in a 2 year fellowship in mouse pathology. With stipends in the range of $60,000 per year, it is recommended that 0.5 M per year be invested for the next 5 years.

b) Development of a Mouse Biology Database to provide access to phenotypic information. Data would be entered from the large published literature, and from current large scale programs. The goal would be a database that is organized on the model of the Mouse Genome Database; it was felt that a large effort would be required to initiate this new database, and initial funding at a level of 4 M per year was recommended.

c) Standardized phenotyping methods will be developed at large scale mutagenesis screening centers. It is recommended that supplemental funds be provided to such centers to enable external investigators to access these facilities for analysis of their mutant mice. An supplement of 5 to 10% of a center budget was recommended for this purpose.

d) Development and Distribution of validated arrays for studying gene expression. Gene expression arrays will provide an important new source of phenotypic information for characterizing new mutants. Significant support should be provided towards making these reagents available to the mouse genetics community.

BREAKOUT GROUP H: Mouse Resources

Cryopreservation

Move toward the broad scale use of cryopreservation for storage and dissemination of strains: Technology development. Investigate the pathogen transmission through sperm as well as the recovery of mice from frozen sperm following artificial insemination (AI), in vitro fertilization (IVF), and intracytoplasmic sperm injection (ICSI). Investigate the effects of strain variation and the effects of mutation(s) [ENU, transgenes, KOs] on recovery. Also, investigate the feasibility of ovary freezing for cryopreservation.

Cost estimate: $1M/lab Two years Two projects

Move to the broad scale use of cryopreservation for storage and dissemination of strains: Outreach. Enable cryopreservation technology to be widely disseminated to the general scientific community. Develop a hands on training course for cryopreservation technology. Also, develop a video/manual describing the details.

Cost estimate: $200K/year 3 training sessions/year/10 students per session

Resource: cryopreserved strains

Develop a storage facility for frozen gametes. Genotype (most/each) strains via gamete analysis. Do extensive quality control on a subset of strains. Develop a cryopreservation laboratory for use by scientists who do not have access to cryopreservation technology. Finally, develop a recovery resource(s) for scientists who do no have access to the technology.

Cost estimate: $1.5M/year 5 years

Resource: Live Strains

Develop a repository(ies) that can accommodate 250 new strains per year. Establish a committee of scientists to evaluate strains for importation. Costs for operating the facility should be paid for from the sale of mice.

Cost estimate: $3M/year 5 years

Database: Disease Models and Phenotypes

Develop a low pass mouse disease models and phenotypes database. Facilitate the expansion of an animal models database that includes not only cancer (already underway) but other disease states and phenotypes. Efforts to make databases interfaceable are imperative (i.e. disease models, molecules, sequences). The initial phase (phase I) should merely provide a "roadmap" or index to other databases containing information about spontaneous models, induced models, etc.

Cost estimate: $300K/year 3 years

Commercial Technology Interactions

For important enabling technologies NIH should explore the feasibility (early on) of establishing a mechanism acceptable to all parties for disseminating the technology(ies) to the general scientific community.

PRIORITY SETTING MEETING FOR MOUSE GENOMICS AND GENETICS RESOURCES

Natcher Building
National Institutes of Health
Bethesda, Md

March 19-21, 1998

INVITED PARTICIPANTS

CO-CHAIRS

NIH ADVISORY COMMITTEE REPRESENTATIVES

NCI Pre-Clinical Models Group

NHGRI Program Planning Subcommittee

Aravinda CHAKRAVARTI, Ph.D., Chair
Co-Chair, Session B: cDNA/ESTs
Department of Genetics
Case Western Reserve University
10900 Euclid Avenue, Room BRB 721
Cleveland, OH 44106
TEL: (216) 368-5847
FAX: (216) 368-5857
E-MAIL: axc39@po.cwru.edu

Charles H. LANGLEY, Ph.D.
Section of Evolution and Ecology
Center for Population Biology
University of California, Davis
Davis, CA 95616
TEL: (916) 752-4085
FAX: (916) 752-1449
E-MAIL: chlangley@ucdavis.edu

Alan R. WILLIAMSON, Ph.D.
760 Lawrence Avenue
Westfield, NJ 07090
TEL: (732) 232-7728
E-MAIL: alan-williamson@home.com

Barbara WOLD, Ph.D.
(See Address Above)

NICHD Advisory Council

Brigid HOGAN, Ph.D.
(See Address Above)

NIGMS Advisory Council

David A. CLAYTON, Ph.D.
Senior Scientific Officer
Howard Hughes Medical Institute
4000 Jones Bridge Road
Chevy Chase, MD 20815
TEL: (301) 215-807
FAX: (301) 215-8828
E-MAIL: claytond@hhmi.org

AGENCY REPRESENTATIVES

Howard Hughes Medical Institute

W. Maxwell COWAN, M.D., Ph.D.
Vice President and Chief Scientific Officer
Howard Hughes Medical Institute
4000 Jones Bridge Road
Chevy Chase, MD 20815
TEL: (301) 215-8803
FAX: (301) 215-8828
David A. CLAYTON, Ph.D.
(See Address Above)

Department of Energy

Ari PATRINOS, Ph.D.
Office of Biological and
Environmental Research
US Department of Energy
19901 Germantown Road, ER-70
Germantown, MD 20874-1290
TEL: (301) 903-3251
FAX: (301) 903-5051
E-MAIL: ari.patrinos@oer.doe.gov

Marvin FRAZIER, Ph.D.
Health Effects and Life Sciences
Research Division
US Department of Energy
19901 Germantown Road, ER-72
Germantown, MD 20874-1290
TEL: (301) 903-5468
FAX: (301) 903-8521
E-MAIL: marvin.frazier@oer.doe.gov

NIH ORGANIZING COMMITTEE

Office of the Director

Harold VARMUS, M.D.
Vida BEAVEN, Ph.D.

National Cancer Institute

Richard KLAUSNER, M.D.
Grace SHEN, Ph.D.
Sue WALDROP

National Human Genome Research Institute

Francis COLLINS, M.D., Ph.D.
Bettie GRAHAM, Ph.D.
Mark GUYER, Ph.D.
Elke JORDAN, Ph.D.
Jane PETERSON, Ph.D.

National Institute of General Medical Sciences

Marvin CASSMAN, Ph.D.
Judith GREENBERG, Ph.D.
Paul WOLFE, Ph.D.

National Institute of Child Health and Human Development

Duane ALEXANDER, M.D.
A. Tyl HEWITT, Ph.D.
Steven KLEIN, Ph.D.
Richard TASCA, Ph.D.
Michael WHALIN, Ph.D.

NIH STAFF ASSISTANTS

Ms. Anita Allen
National Human Genome Research Institute

Ms. Stephanie Walker
National Human Genome Research Institute

REPORT FINALIZED JUNE 11, 1998