Working Group Meeting on Studies of Large-Scale Variation

Location:
NIGMS, NIH

Start Date: 6/20/2001 8:00 AM

End Date: 6/20/2001 2:30 PM

The NIGMS working group was charged with identifying research trends and opportunities related to large-scale patterns of genetic variation. The past few decades of research focusing largely on reductionism have yielded vast amounts of data. In addition, the various genome sequencing projects, as well as structural and functional genomics initiatives, are producing data far more rapidly than scientists can analyze them and understand their implications to biology and to health. Indeed, as the working group pointed out, as complex and overwhelming as the current data are, they are only the beginning. Every protein structure, every DNA sequence, every gene expression pattern has the potential of varying among individuals, among species, among populations within a species, and across environments. It will soon be possible to utilize information on thousands of variable genetic sites to investigate the relationships among genotypes, phenotypes, and environments. Studies of genetic variation are in their infancy, and it is vital to encourage research in this area.

There is no single way to characterize large-scale variation. One approach is to consider the combinatorial effects of many variable sites, whether the scale is within a gene or across a genome. Comparative genomics, where the goal is to identify patterns of variation among genomes, is also a productive way of identifying attributes of variation, such as which genomic regions are rapidly evolving. It is also important to study the context or environments in which genetic variation arises, is selected, and is maintained. Finally, variation occurs at every biological level of organization, from DNA sequence to protein structure to metabolic pathways to cell dynamics to individual phenotype to population characteristics.

Future Research Areas

Biological variation underlies differences in the expression of many human traits, including genetic disorders, responsiveness to drugs, susceptibility and resistance to infection, and response to trauma. Genetic and environmental variation may be related to the onset of a disease, its specific symptoms, and its severity, even if the disease-causing gene itself is not directly affected. The following areas of research may uncover important features of biological variation.

Mid-Level Reality

Studies of genetic architecture have historically focused on associations of genotype and phenotype (e.g., between DNA markers and a disease). In the future, it will also be important to focus on many levels of biological processes between DNA and a phenotype. A significant NIGMS effort focuses on the detailed analysis of the physical, biochemical, and genetic pathways that direct cellular development and metabolism. A next step will be studying how these complex systems diverge in different states (e.g., in a single individual, among healthy individuals, in the presence of disease, in different organisms) and to address how these complex systems evolve.

Context Dependency

How genes are expressed depends on their cellular, developmental, physiological, and environmental context. The importance of context dependence is particularly important in medical research because study populations are rarely a random sample of the general population. For example, certain mutations in BRCA1 are predictive of breast cancer in 52 percent of the cases in which four or more family members have been affected. The same mutations are associated with only 1 percent of the sporadic cases of breast cancer. Clearly, BRCA1 operates in a larger context about which we know little. We need better tools and models for identifying important contextual features and determining how they interact.

Evolution of Genome Properties

An emerging area of research focuses on such issues as the evolution of haplotypes, selection for increased genetic interactions, and the evolution of recombination and methylation patterns. For example, rates of crossing-over vary within a genome; the Y chromosome rarely, if ever, recombines, but there are other "hotspots" of recombination. Recombination rates also vary dramatically among species; two closely related species of Drosophila, for example, differ 14-fold in their overall recombination rates. Such differences have important consequences for genome organization and evolution. The organization and evolution of haplotypes is an emerging area of research. For example, studies of the human Y chromosome and mitochondrial DNA have led to a good understanding of early human history and migration--a necessary prerequisite to studying the genetics of specific populations at risk for disease. We also know that specific haplotypes, while not causal, are associated with such diseases as multiple sclerosis, diabetes, arthritis, Parkinson's disease, and AIDS.

Extensions to Other Organisms

Many organisms have been studied for their value in agriculture or ecology. Thus there is considerable information about the population structure, natural history, and genetics of these systems. It will be valuable to take advantage of this wealth of information to study variation in the natural settings in which it evolved.

Bioinformatics

The study of biological variation depends heavily on rich data sets; researchers need the ability to access many kinds of information (e.g., DNA sequence, protein structure, development, natural history, and phenotype) in organisms from different habitats, from different populations, or from different species. In addition, whether one is looking at a clinical phenotype or at a gene expression array, it is important that experts agree on data standards so that data can be compared, combined, and shared. Organizing, maintaining, and making this information available should be included in any initiative.

Improved Dynamic Modeling and Statistical Methods

Mathematical approaches to studying biological variation have changed little in several decades. The working group pointed out that there is a need to develop new dynamic models to illuminate how systems interact. Just as important, it is critical to study the nature of biological and mathematical assumptions on models and statistics. For example, assumptions about cell biology or protein structure (e.g., that synonymous changes are not under selection or that mutation rates are the same throughout the genome) may seriously affect the validity of our analyses and predictions.

Recommendations

Studies of DNA sequences, protein structures, and basic cell processes have provided and will continue to provide rich data for studies of biological systems. The scientific scope of the proposed initiative would include:

  • Normal variation in molecular, cellular, and developmental processes;
  • Interactions among genes and relevant components of the environment;
  • Evolution of genome properties;
  • Use of a wider range of model systems with the goal of studying variation in its evolutionary context;
  • Mathematical models for studying biological properties of variation;
  • Statistical tools based on biologically relevant assumptions; and
  • Improved datasets and improved tools for using them.

The working group felt strongly that there should be opportunities both for individual research programs and for large, collaborative efforts. Based on the discussion at the meeting, NIGMS staff recommend a two-fold approach to addressing these research areas:

  • Rewrite and reissue the current announcement "Genetic Architecture of Complex Phenotypes" to bring it up to date and include research described by the working group. This will provide opportunities for individual investigators.
  • Explicitly include studies of genetic variation in the next release of the BISTI (Biomedical Information Science and Technology Initiative) Centers announcement. This will allow for collaborative efforts and has the particular advantage of putting studies of genetic variation in the context of other high-priority initiatives.

The working group also raised the issue of explicitly including both minority scientists and minority-serving institutions. The rationale is two-fold. First, to be credible, studies of human variation should be designed and conducted by a diverse and knowledgeable group of scientists. Second, in many areas of cutting-edge science, such as genomics and evolutionary biology, minority scientists are seriously underrepresented. It is important to establish programs to recruit and train those scientists now. The working group stressed that real partnerships among scientists and institutions are vital to addressing this goal. Collaborations that address training and infrastructure needs in minority-serving institutions are especially valuable.

Working Group Members

Alan Templeton, Chair
Washington University
One Brookings Drive
St. Louis, MO 63130
(314) 935-6868

Andrew Clark
Pennsylvania State University
326 Mueller Lab
University Park, PA 16802
(814) 863-3891

Anna DiRienzo
Department of Human Genetics
University of Chicago
920 E. 58th Street
CLSC 507F
Chicago, IL 60637
(773) 834-1037

Marc Feldman
Stanford University
HERRIN 478A
Stanford, CA 94305-5020
(650) 725-1867

Ira Herskowitz
Department of Biochemistry and Biophysics
University of California, San Francisco
Box 0448, HSE 901H
San Francisco, CA 94143
(415) 476-4977

David Kingsley
HHMI and Department of Developmental Biology
Stanford University School of Medicine
Beckman Center B300
279 Campus Drive
Stanford, CA 94305-53295
(650) 725-5954

Rick Kittles
National Human Genome Center
Howard University
2041 Georgia Avinue
Room 615
Washington, DC 20060
(202) 806-6979

Richard Lewontin
Department of Organismic and Evolutionary Biology
Museum of Comparative Zoology Labs 317b
26 Oxford Street
Harvard University
Cambridge, MA 02138
(617) 495-2419

Maria Fatima Lima
School of Graduate Studies and Research
Meharry Medical College
1005 D.B. Todd Boulevard
Nashville, TN 37208-3599
(615) 327-6533

Jean MacCluer
Department of Genetics
Southwest Foundation for Biomedical Research
PO Box 760549
San Antonio, TX 78245-0549
(210) 258-9490

Andrew Murray
Department of Molecular & Cellular Biology
Harvard University
16 Divinity Avenue, Room 3000
Cambridge, MA 02138-2097
(617) 496-1350

Andrej Sali
Rockefeller University
1230 York Avenue
New York, NY 10021
(212) 327-7550

Oliver Smithies
University of North Carolina
Chapel Hill, NC 27599
(919) 966-6913

Roland Somogyi
Molecular Mining Corp
128 Ontario Street
Kingston, Ontario
K7L 2Y4
(613) 547-9752