By Tom Ulrich
Detecting genetic variants has never been easier, thanks to cheaper, faster sequencingand constant improvements in analytical methods. And as more people have their genomes sequenced, researchers are finding variants by the millions. The 1000 Genomes Project (one of the first efforts to document the scope of human genetic variation), for instance, identified more than 88 million variants across the human genome.
Interpreting a variant's impact, however, and how it contributes to a trait or disease remains a challenge, especially in the clinic. Testing laboratories sometimes disagree in how they interpret genetic findings, with one flagging a gene variant as "likely pathogenic," and another calling it "uncertain." This stems from a historic lack of standards for interpreting variants' effects, and of consensus around what those standards might entail.
In recent years, researchers and clinicians behind two National Institutes of Health-funded efforts, the Clinical Genome Resource (ClinGen, a resource for defining genes and variants' clinical relevance) and ClinVar (a public archive of variant association reports), have worked together to develop resources that geneticists and genetic counselors can use to interpret a patient's results, and to contribute new findings as well.
ClinGen has also worked to create and rally the community behind consensus standards for interpreting variants, and to help resolve conflicting interpretations of data. And working with organizations like the Global Alliance for Genomics and Health (GA4GH), ClinGen has developed standards and policies for sharing data around the world while protecting individuals' privacy.
ClinGen and ClinVar collaborators recently published 25 papers in a special issue of Human Mutation outlining both projects' progress over the last five years. Institute member Heidi Rehm — who leads the ClinGen team at the Broad and serves as medical director of the institute's Clinical Research Sequencing Platform and co-director of the Broad Center for Mendelian Genomics — was an editor of the special issue and corresponding author on an accompanying editorial. She talked to the BroadMinded Blog about the two efforts' impacts to date, how far variant interpretation has come, and what she thinks the future holds.
Why is standardizing variant interpretation so difficult?
Rehm: For most variants, we have only bits and pieces of data, sitting in different laboratories around the world. Getting access to those data and deciding how to combine and weight the evidence is challenging.
Historically, what have been the best practices for clinical geneticists for interpreting a patient's genetic findings?
Rehm: Observing the variant in affected individuals and showing its absence in healthy individuals has been the most common type of evidence. However, it is only recently that we have had access to large population databases like the Exome Aggregation Consortium (ExAC) and Genome Aggregation Database (gnomAD) browsers to better define the population frequency of variants. We also now realize that even if a variant found in a sick patient is absent from all others, it does not mean it's pathogenic (disease-causing).
What would you say have been ClinGen and ClinVars' biggest accomplishments over the last five years?
Rehm: I would say that our biggest accomplishment to date has been to work with laboratories to help them share their variant interpretations in ClinVar, and then catalyze collaborative efforts between laboratories and with our Expert Panels to help resolve differences in interpretation and combine evidence to clarify the significance of variants.
How do ClinGen, ClinVar, and other data sharing efforts like ExAC, gnomAD, GA4GH, or the Matchmaker Exchange, complement each other?
Rehm: ExAC and gnomAD are critical to supporting variant interpretation, particularly for ruling out a pathogenic effect if a variant is found at a frequency higher than that of the disease it is associated with. GA4GH helps build standards for how we share data, ensuring that the global community can most effectively share and benefit from data being generated around the world. The Matchmaker Exchange is a platform to build evidence for a gene’s role in a disease, a critical step for ClinGen’s efforts to define which genes have valid evidence to be called causal.
What do you think are the biggest challenges for ClinGen and ClinVar, and for clinical interpretation today?
Rehm: We still don’t have efficient access to genetic and clinical data in ways that we can readily use for clinical interpretation of rare variants. Today we figure out who may have data on a variant because they submitted an interpretation to ClinVar, but then we have to contact that lab and ask them to dig up the information and share it to help interpret the variant. Even if data are published in the public domain, they are often embedded in text and difficult to extract efficiently.
We need to build platforms that allow robust and protected sharing of genetic and health data, as well as ways to store and share biological evidence that conform to interoperable standards that are widely used by the global community. We also need to have more diversity in the sources of data, ensuring access to evidence from all populations.
How about the biggest opportunities?
Rehm: We are generating enormous datasets from many countries where genetic data is linked to health data, such as the UK Biobank, All of Us Research Program, and FinnGen. If we can effectively leverage and share those data, we will be able to continuously learn from the data and apply them to improve the health of individuals with disease or at risk for disease.
How can geneticists, clinical laboratories, and genetic counselors best help build consensus in the field?
Rehm: By sharing their variant interpretations in ClinVar, participating in ClinGen’s Expert Panels to develop better guidance for the interpretation of genomic information, and applying those methods to genes and variants discovered through clinical testing and research studies.
ClinGen is supported by the National Human Genome Research Institute with additional support from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. ClinVar is supported by the National Center for Biotechnology Information.
Rehm HL, Berg JS, Plon SE. ClinGen and ClinVar: Enabling genomics in precision medicine. Human Mutation. Online October 12, 2018. DOI: 10.1002/humu.23654.