| <html> |
| <body> |
| <p><strong>CONCEPT CLEARANCE FOR RFA</strong></p> |
| <p><strong> Rapid Investigation of Genetic Associations in Population Studies </strong></p> |
| <p><strong>NHGRI Advisory Council, May 2007 </strong></p> |
| <p><strong>Purpose</strong></p> |
| <p>This RFA is proposed to support the investigation, in well-characterized population samples, of genetic variants identified as potentially causally associated with complex diseases in genome-wide association (GWA) and other genetic studies. The goal of this initiative is to utilize existing prospective cohort studies and clinical trials, and the investigators most familiar with their intricacies, to: 1) determine the population impact of putative risk variants, including prevalence, disease risk, and associations with other health characteristics; 2) identify modifiers of gene-trait associations, particularly those related to lifestyle factors or medication use; and 3) identify potential clues to gene function, by examining associations of putative risk variants with related phenotypic characteristics such as laboratory measures or imaging findings.</p> |
| <p><strong>Background </strong></p> |
| <p>Large-scale GWA studies are a crucial first step in the identification of genetic variants related to complex diseases, but they are only a first step. Several subsequent phases of investigation are needed to eliminate potentially spurious associations, characterize the full spectrum of variants in an associated region, determine the impact of variants on gene function, and design appropriate intervention strategies. Many, but not all, of these steps are already underway or in planning to follow up current or proposed GWA studies. Most proposals for initial GWA scans, for example, now include plans for repeating an entire genome scan, or a subset of selected variants, in independent replication studies. Fine mapping of associated regions, and identifying and typing rarer variants found by targeted sequencing, are increasingly being pursued in initial or replication population samples. Once a small number of repeatedly associated, potentially causal variants is isolated, however, investigation often shifts away from human populations to the laboratory, to examine the impact of variants on gene function experimentally in model systems.</p> |
| <p>Yet much remains to be learned from well-characterized human population samples in which potentially causative variants have been, or could be, assayed. Important clues to gene function could be identified by examining associations with related or intermediate phenotypes such as hormone levels or bone density. More importantly, the potential population impact of variants of interest is often poorly described by the often highly-selected or otherwise non-representative group of cases in whom they were initially identified. The “epidemiologic architecture” of these putative causal variants—their population prevalence; prevalence in race/ethnic subgroups; relative risk of rigorously-defined, incident disease; consistency of association across subgroups defined by age, sex, race/ethnicity, or exposures; and potential modifiability of associated risk—is rarely pursued in population studies once a genotype-phenotype association is identified. This information is critical, however, to determining the health implications of a given variant and the priority it should receive for identifying and testing interventions to reduce its associated risk. This information may also be quite valuable in exploring gene function, since the epidemiologic approach of genetic investigation, starting from observed phenotypic characteristics and moving more proximally to gene pathways and sequence variants, complements well the laboratory approach of moving from DNA sequence to function to phenotype.</p> |
| <p>Reasons for failing to pursue population-based investigations of genetic variants further may include lack of appreciation of their potential for illuminating the population importance and physiologic effects of gene variants, lack of information on available population studies and the data they contain, lack of familiarity with the intricacies of these databases and the requisite analytic methods, lack of access to the data, and lack of patience with the potentially arduous process of examining the full range of associations and interactions in population samples. Although these obstacles might be addressed by cataloguing potentially available studies and encouraging collaboration, the complexity of these datasets is such that the epidemiologic investigators who develop and routinely work with them remain uniquely positioned to carry out these analyses. Rarely, however, are they supported to do so. Although no single group of investigators can fully mine the massive potential of these large-scale datasets, and consequently the wide sharing of de-identified data continues to be strongly encouraged, not all studies are suitable for widespread sharing of genotype-phenotype data on individual participants. Widespread sharing of association data should present far fewer obstacles.</p> |
| <p>More importantly, defining the risk associated with a specific genotype is essentially an epidemiologic problem, similar to characterizing any other risk factor, and requiring a detailed understanding of bias, confounding, and interaction. Such investigations may best be undertaken by experienced epidemiologists using data from the large-scale, prospective, population-based studies they have designed and carried out, because of the complexity of these datasets and the potential biases involved in exposure assessment and follow-up. Clinical trials, particularly primary prevention trials, may provide similarly representative and well-characterized population samples, with the added value of randomized interventions that may suggest potential avenues for modifying genotype-phenotype associations.</p> |
| <p><strong>Research Scope and Objectives</strong></p> |
| <p>This RFA would support multiple investigative groups with population-based epidemiologic and analytic expertise, and affiliated with existing population studies, to investigate and disseminate information on the epidemiologic architecture of putative disease-associated genetic variants. These groups would work together to develop criteria and methods for identifying highly-replicated, putative causal variants suitable for epidemiologic pursuit within individual studies. Vigorous and productive debate will likely ensue on the evidence needed to justify extensive population-based characterization of a variant (such as the number of independent replications, number of subjects and consistency of associations, number or type of traits associated), which could provide guidance to the studies participating in this program and to population studies in general. It will be important to balance the advantages of group consideration and prioritization against the speed and flexibility of pursuit possible within an individual study, so that research would neither be unnecessarily slowed nor undesirably duplicated. </p> |
| <p>Subsequent investigation of variants identified as suitable for epidemiologic characterization in individual studies supported by this program might include: 1) identifying appropriate study participants to be genotyped and relevant phenotypic and exposure data to be used in analyses; 2) retrieving existing samples, performing whole genome amplification and/or transformation of cryopreserved cells if needed, and isolating, quantifying, and aliquoting DNA samples if not already funded to do so; 3) conducting genotypic and other evolving genomic assays (such as gene expression in transformed lymphoblasts) as needed to characterize the reported association; 4) analyzing the resulting data in detail, in relation to all or nearly all the relevant phenotypic and covariate/exposure data available, and synthesizing the results into a readily comprehensible form; and 5) disseminating the findings in user-friendly formats through standard databases such as dbGaP. The most rapid possible timeframe for these steps should be proposed, and should continue to accelerate over the course of the program as efficiencies and best practices evolve. </p> |
| <p>A key step in this process will be fully utilizing the available data outside the primary genotype-phenotype association in ways that maximize information on population impact and potential gene function, while minimizing spurious findings and difficulty sifting through the resulting analyses. Innovative bioinformatic approaches for identifying and displaying potentially important findings will be encouraged to increase the ease of use of the association results.. Sharing of expertise and experience in this and other areas across investigative groups will be a major goal of this collaborative program, with the intent of developing best practices for rapid investigation of putative genetic associations and making the resulting findings available and understandable to investigators from a wide range of genomics-related disciplines.</p> |
| <p>Early in the project period, investigative groups and a Coordinating Center supported by this RFA would form a Steering Committee to examine the phenotypic, covariate, and exposure data available in each participating population study, including the variables collected, methods used, and documentation available. Quality and quantity of DNA and related genomic and other samples would be assessed and documented in a standard format across all participating studies. The Steering Committee would identify strengths and weaknesses of the various studies in terms of traits and covariates available for investigation, and would suggest best uses for specific studies, such as examining the impact of a particular intervention on a given genotype-phenotype association, or the risk of a specific trait associated with a given genetic variant. The Coordinating Center would catalog all this information across studies and disseminate it through a program website, standard databases, and other means. Approaches to analysis would be shared and reviewed in periodic investigators’ meetings, using summary data derived from the studies themselves. Investigators would be required to propose a plan for rapid release of analytic results, in both a summary, user-friendly form and a more detailed, comprehensive form, to be compiled by the Coordinating Center and provided through databases such as dbGaP. Appropriate embargoes on submission of publications from these findings would be applied program-wide. Sharing of genotype or phenotype data on individual study participants across studies or outside the program would not be necessary to participate in the program, but would be strongly encouraged in keeping with NHGRI and NIH policies.</p> |
| <p>Each applicant would be asked to propose a method for rapid investigation and dissemination of genetic association findings in a cohort study or clinical trial, or consortium of cohorts or trials, involving large numbers (10,000 or more as a rough estimate) of well-characterized participants in whom extensive phenotype, covariate, and exposure data are available and high-quality DNA is isolated and accessible, or could be isolated from stored specimens. Studies should be of sufficient size and breadth to permit examination of a wide range of variants and traits, and to identify associations and interactions with sufficient power. Neither recruitment of participants nor collection of phenotype, covariate or exposure data, or biospecimens would be supported by this initiative. Whole-genome amplification and/or isolation of DNA in a high-priority subset of subjects could be supported as funding permits. Applicants would be encouraged to include in each study a broad range of participants approximating the diversity of the U.S. population on factors such as age, sex, race/ethnicity, socioeconomic status, US geographic region, and urban/rural residence. Studies involving under-represented racial/ethnic minorities, particularly those suffering disproportionate burdens of disease, would particularly be sought. Selection criteria for funding would include studies that provide a diversity of trait and exposure information of population health importance, and a diversity of population groups (particularly U.S. racial/ethnic minorities), to the program as a whole. </p> |
| <p><strong>Mechanism of Support</strong></p> |
| <p>This initiative would use the NIH U01 (Cooperative Agreement) award mechanism. Three to five investigative groups, each representing a single population study or a consortium of studies, and each investigative group involving roughly 10,000 or more total participants, would be selected for award. A separate solicitation for a Coordinating Center would also be released. </p> |
| <p><strong>Funds Available</strong></p> |
| <p>NHGRI will commit approximately $32M over a four-year period to support these cooperative agreements; roughly $10M of these costs are for genotyping (and whole genome amplification, cell transformation, and DNA isolation as needed) on roughly 40,000 or more participants. </p> |
| <p>PAGE </p> |
| <p>PAGE 2</p> |
| </body> |
| </html> |
|
|
