Sequencing studies, including the 1000 Genomes Project and many disease studies, have identified more than 100 million sites in the human genome that vary among people. However, it is likely that most of the variants do not contribute to differences in function, traits, disease risk, or clinical outcomes. Thus, methods to determine which variants or sets of variants causally lead to differences in function are needed for a deeper understanding of the processes that lead to differences among people in many traits, including disease risk and clinical phenotypes.
For this Funding Opportunity Announcement (FOA), the term “disease or trait” is used broadly, to encompass diseases, risk of diseases, protective effects against diseases, molecular phenotypes, organismal phenotypes, clinical phenotypes or outcomes, traits, responses to therapeutic drugs or vaccines, and other outcomes relevant to human health and disease.
Genome-wide association studies (GWAS) have been highly successful in looking at genome-wide sets of genetic variants to detect their associations with disease and many other traits. The GWAS catalog provides a comprehensive listing of almost 60,000 SNP-trait associations (https://www.ebi.ac.uk/gwas/). Whole-exome and whole-genome sequencing studies have identified rare variants associated with disease.
The linkage disequilibrium patterns in the genome mean that disease- or trait-associated regions may contain many genes, other functional genomic elements, and variants. Thus, determining which variants affect the function of genes or other genomic elements is a major challenge. Even for situations with a higher likelihood of a variant being causal, such as a missense variant in a protein-coding region, the functional significance of this variant generally cannot be determined with available data. When variants in known disease- or trait-associated regions of the genome are studied, it can be unclear how they contribute to phenotype, without, for example, replicating the phenotype in a model system. In GWAS and whole-genome sequencing projects, the majority of disease- and trait-associated variants fall outside the coding regions and usually cannot be interpreted without substantial functional data.
Understanding which variants in a person’s genome lead to disease risk and changes in health is important for clinical implementation supported by commercial entities. Genomic sequencing is increasingly being performed in clinical settings and precision medicine, and function data can be essential for classifying the variants identified in patients. Linking high-throughput assays of variant effects to patients’ phenotypic characteristics and treatment responses could increase our understanding of the functional consequences of variants. The Genomic Medicine Meeting IX (https://www.ncbi.nlm.nih.gov/pubmed/28340351) discussed several issues related to using functional data to understand clinical genomes. The recommendations included improving approaches for studies of animal and cell models to better understand genetic variants observed in human studies, improving the connection of functional data with standardized phenotypic data, and enhancing collaborations between clinical and basic scientists. It will be important to determine what types of functional data would be most valuable for interpreting genomic variation in clinical settings at the individual and population levels, and how these data could best be used to advance the science of clinical medicine.
Many resources help the interpretation of variants identified through association studies and clinical sequencing. For example, the GTEx Project (https://www.gtexportal.org/home/) studied the effect of variants on gene expression across many tissues from many individuals. The ENCODE Project (www.encodeproject.org), the NIH Roadmap Epigenomics Program (https://commonfund.nih.gov/epigenomics), and other genomics projects have mapped many types of functional genomic elements across the genome. The ClinGen Project (https://www.clinicalgenome.org/) is building an authoritative central resource that defines the clinical relevance of genes and variants for precision medicine and research. The eMERGE Network (https://emerge.mc.vanderbilt.edu/) is sequencing and assessing the phenotypic effects of ~100 clinically relevant genes, including studying the penetrance of rare variants.
Commercial research efforts are needed to develop novel and transformative approaches for the next steps of identifying which variants, genes, and other genomic elements cause disease or result in other traits, and how they function to do so. Even variants in relevant functional genomic elements may not cause differences in function. A major need is to figure out what types of data, methods, and approaches will allow researchers to discover which genes or other genomic elements contribute to disease or traits, how variants in those elements result in higher or lower risk, and how they affect disease and biological processes. Differences in function may be at the molecular or cellular level, leading to differences at the tissue and organismal level. Generalizable approaches are needed that provide a deeper understanding and go beyond association to investigate the black box that currently exists between variants and diseases and other traits.
This FOA aims to support commercial research that develops novel, transformative, and generalizable genomic approaches to study the functional and disease effects of genomic variation, specifically how differences in sequence lead to differences in genome function, and to better understand how functional differences lead to disease risk or traits, or how to this knowledge can be used clinically. These new approaches could fall into a range of activities, from exploring novel concepts, developing new methods, or developing new ways to analyze data that will substantially advance the ability to understand the functional consequences of sequence variation and provide fundamental knowledge to directly or indirectly accelerate scientific and medical breakthroughs that improve human health.
This FOA aims to develop commercial approaches and tools that can be used broadly to study the relationships among genetic variation, function, traits, and disease. The focus should be on developing commercial approaches that can be applied generally across multiple disease or trait outcomes. Approaches may be tested using specific genomic elements, variants, cell types, diseases, traits, or model organisms; however, the generalizability of the approach must be explained well. Approaches that are comprehensive across the genome and assay many or all variants at once are encouraged. This PA is not intended for approaches that are only applicable to, or tailored towards, study of an individual disease, gene, or variant.
Studies that examine how genetic variants affect molecular phenotypes (e.g., transcriptome and epigenetic profiles) are appropriate. Studies that look at how genetic variants affect tissue or physiological function may be appropriate if the approaches can be generalized across variants, tissues, diseases, or function. Some approaches may relate to the choice of samples or biological contexts to be studied; for example, studying functional data types in people prior to and during the time that they are developing a disease may be informative, as would studying people who are at high risk of a disease but have not developed it. Proposed studies could also investigate genetic interactions, looking at combinatorial effects of multiple variants. The application should explain how generalizable, broadly useful, and transformative the approaches will be.
Approaches may include one or more of these aspects of genomics:
Functional genomics: Applications may propose to develop new commercial functional genomic approaches to study how differences in sequence lead to differences in function, and how this informs disease or related biological processes. This may be for specific types of variants, genomic elements, or genomic regions, for regulatory interactions, or for networks. Use of model organisms or model systems is encouraged, especially where approaches are especially suited for these systems; however, information forthcoming from these models should be relevant to humans, e.g., through understanding of general principles or functional consequences of variants in conserved sequences, or methodological advances. Examples include developing novel gene editing approaches in cell lines or model organisms or developing methods to study transcription factor binding differences and their effects on transcription and cellular phenotypes, insertion or deletion variants that affect nucleosome spacing and regulation, or variants that cause differences in epigenomic marks and their effects on disease risk. Methods are encouraged that can help to infer causality or indicate which molecular function differences result in phenotypic differences.
Technology development: Applications may propose to develop novel commercial genomic technologies that assay how sequence differences in functional genomic elements lead to differences in function and phenotypes. Development of comprehensive, genome-wide, or high-throughput methods for commercial availability are encouraged. Examples include new assays for function that can detect differences among variants in genomic elements, or improved sequence editing methods that allow single or multiple variants in functional elements to be studied, preferably for many functional elements at a time.
Clinical genomics: Applications may propose to develop new commercial approaches that use functional genomics data to better interpret genomes in a clinical context. This may include developing data sets to deduce whether variants are pathogenic or benign, developing methods to use functional data to better assess penetrance or predict individuals at risk of disease, developing approaches to connect in a comprehensive way phenotype data with functional data for sets of variants in the context of precision medicine and health, or other ways to interpret VUS. Applications may also propose to commercially implement functional data or functional interpretation methods in clinical settings.
Data integration and analysis: Applications may include the commercial development of novel or significant improvement of existing computational methods to integrate functional data and other data types to help to understand which variants are causal for phenotypes, and how variants lead to differences in function in relation to molecular or cellular phenotypes, clinical phenotypes, disease risk or mechanism, other traits, or other ways of interpreting clinical genomes. Applications may propose to study which data types are most informative for understanding how variants affect function or disease risk or may be interpreted clinically.
Examples of possible commercial research topics are:
- Developing assays for relating genetic variation to function. These could be assays of how molecular functions differ based on variants in functional genomic elements.
- Developing tools that interpret the function differences and pathogenicity of variants. This could be based on integrating various functional and population data types with phenotype data.
- Providing reports of how genetic variants function differently as a service that might include lookup of variant functions in a database or be based on assays.
- Developing tools that narrow the set of causal variants for a phenotype by integrating functional data sets. For example, including tissue expression data could help limit the set to those in tissues related to a disease.
- Providing risk estimates by using data sets that integrate genetic information across the genome.
This FOA aims to support the development of novel commercial approaches in this area, so applications whose major aim is the production of functional data sets using existing or established approaches are not appropriate for this FOA; NHGRI program officers may be able to suggest other types of support for those activities. A major outcome of this program will be results showing which data types would be most broadly useful for understanding how genetic variation leads to differences in the function of genomic elements that lead to differences in disease risk, traits, and pathogenicity. Future funding announcements may be developed to scale up the production of such data sets.
Genomic studies have generally lacked inclusion of substantial numbers of non-European-ancestry participants, so the production or analysis of data from individuals of diverse races, ethnicities, and ancestries is encouraged.
NHGRI encourages investigators who plan to collect phenotype or environmental exposure data to use the standard protocols in the PhenX Toolkit (www.phenxtoolkit.org).
Projects that focus on tumor or somatic cancer genomics will not be appropriate for NHGRI funding.
To enable the widest use of the functional data, any human subject consents should allow broad, general research use of the data, with no restrictions on the types of researchers who may use the data. Where possible, use of samples consented for submission of data into unrestricted databases is strongly encouraged.
In addition to this PA, NHGRI participates in several funding opportunities https://www.genome.gov/10000991/nhgri-funding-opportunities-research/, including the parent R01 announcements. Under those parent announcements, NHGRI will support the development of resources, methods, and technologies that will accelerate research in understanding the structure of genomes, understanding the biology of genomes, understanding the biology of disease, advancing the science of medicine, and improving the effectiveness of healthcare. NHGRI will also support research in several cross-cutting areas, including the ethical, legal and societal implications of genomics research, clinical implementation of genomics, bioinformatics, technology development, and research training and career development. Although clinical trials are not allowed under this FOA, NHGRI supports such research under parent NIH FOAs that allow them.
Potential applicants are strongly encouraged to contact NHGRI staff early in the application development process.
See Section VIII. Other Information for award authorities and regulations.