NEW YORK – A team of UK investigators has proposed new guidelines for assessing the significance of variants in noncoding regions of the genome, providing a platform for better interpreting their role in rare disease.
The researchers, led by scientists at Genomics England, the University of Manchester, and the University of Oxford, outlined their recommendations in the journal Genome Medicine this week. They come at a time of a shift in the field toward clinical whole-genome sequencing and away from targeted assays and whole-exome sequencing. This has made the need for recommendations on how to cope with variants of unknown significance more pressing.
There is also demand to make the most of whole-genome sequencing to produce a diagnosis for patients and families. According to Jamie Ellingford, lead genomic data scientist in rare disease at Genomics England and a research fellow at the University of Manchester, it was this desire to improve diagnostic outcomes that was at the heart of the effort to devise the guidelines.
Ellingford, lead author on the publication, said the work emerged from a collaboration with Nicola Whiffin, a group leader at the University of Oxford and the study's corresponding author.
"In early 2020, we decided that there was clearly a gap, and large inconsistency, in how noncoding regions of the genome were analyzed in clinical labs," said Ellingford. "We recognized that given our different areas of expertise, but common motivation, we could spearhead some of these developments."
To accomplish this, the authors assembled an expert team of nine scientists and clinicians from major genomics laboratories in the UK, including four of the NHS Genomic Laboratory Hubs that serve its Genomic Medicine Service. According to the paper, the panel consisted of experts with "wide-ranging expertise in clinical variant interpretation, with specific experience in variants within noncoding regions."
As also noted in the paper, clinical whole-genome sequencing has become a first-line test for patients with undiagnosed conditions, but the approach yields variants in noncoding regions, the significance of which is unknown. Currently, labs look to a 2015 guidance from the American College of Medical Genetics and Genomics and the Association for Molecular Pathology for how to interpret clinical sequencing findings, but the seven-year-old guidance is more focused on protein-coding regions. However, the authors found that the rules laid out in the 2015 guidelines could also be applied to noncoding regions. For nine of the ACMG/AMP rules, the authors did propose some specific modifications around prediction tools, often to make them more conservative by downgrading the strength of evidence supporting a pathogenic classification.
They also made recommendations on how to define and filter candidate noncoding regions, stating that labs should only clinically interpret variants that fall into regulatory elements that have well-established or functionally validated links to target genes, as well as genes that have documented associations to the phenotype of interest. Such an approach is necessary to avoid a "huge burden of interpretation" on labs faced with scores of variants of unknown significance.
According to the paper, the draft recommendations were tested out by a group of clinical scientists using test variants and refined by a group of experts from the UK, the Broad Institute in the US, and the Garvan Institute and Murdoch Children's Research Institute in Australia.
Ellingford underscored the value of the Genomics England dataset in drafting the recommendations. He said that one reason such guidelines have not yet been created is that the data necessary to analyze the noncoding genome en masse hasn't been routinely available in diagnostics. He said that the 100,000 Genomes Project and associated initiatives in the UK catalyzed the development of the Genomic Medicine Service within the UK National Health Service. "Now the challenge is a reality that clinical scientists face every day," he said.
As such, the availability of recommendations comes "just right and at the time" when the community could benefit from having a standardized way to tackle issues around interpreting variants found in noncoding regions, he added.
The authors have already discussed and invited feedback on the guidance at the tail end of 2021 at the Association for Clinical Genomic Science meeting, which was held virtually. They will also give an overview of the recommendations at the American Society of Human Genetics annual meeting, which will be held in Los Angeles in October.
Ellingford conceded that interpreting variants in noncoding regions is a "challenging area" and that the tools employed for interpreting their impact are designed based on just a handful of pathogenic variants known to exist in such regions. In addition, the same regions involved in regulation, for example, can be "both temporally and spatially specific." As new pathogenic variants and regulatory elements are discovered and characterized, he said, guidance on how to interpret them will also evolve and develop.
While the guidelines may involve over time, it is very much the intention of the authors that they be adopted and implemented. At ASHG, Ellingford said, he and other authors aim to present a tutorial on accessing and using evidence to classify noncoding variants.
"We hope that this will encourage the community to adopt them in their everyday work, and in doing so, put them under greater scrutiny and thereby facilitate their development," said Ellingford. "As we test these guidelines with more and more data, and openly share the variants and their classifications, we will be able to better use them routinely in a healthcare setting."
Heidi Rehm, a coauthor on the paper and codirector of medical and population genetics at the Broad, said that she expects the recommendations will be "very useful for the community" and that the next version of ACMG/AMP standards, actively under development, will "use the framework as a starting point for providing more explicit guidance for evaluating noncoding variation."
Rehm, who was not a member of the expert panel that drafted the recommendations, said that all clinical labs should read them, and that a focus should be on unsolved cases with suspected recessive genetic disease for which only a single heterozygous pathogenic variant has been identified in a gene. "The other variant is likely a noncoding regulatory variant," she said.
According to Rehm, the community is hungry for more guidance at the moment, especially as clinical labs in the past several years have trended towards adopting whole-genome and whole-exome sequencing instead of favoring the targeted sequencing approaches they once did.
For example, in May, the European Society of Human Genetics updated its guidance for clinical whole-genome sequencing, providing 44 recommendations covering everything from general implementation of whole-genome sequencing to diagnostic routing, bioinformatics, quality assessment, ethical considerations, and reporting.
Rehm noted that there are multiple other efforts to shape guidelines for the community. Gene-specific guidance is being developed by ClinGen expert panels, for instance, and ClinGen's sequence variant interpretation working group has also published recommendations.
In a recent, separate paper in Nature Medicine, Rehm called for greater global sharing of data to improve diagnosis of rare disease via sequencing. In the paper, she noted that interpretation of noncoding variation is emerging and that addressing noncoding variation will require larger datasets and more cost-effective and scalable methods for data interpretation.
According to Ellingford, the new recommendations are a step forward in addressing noncoding variation. "We need to be screening these regions, but interpretation is difficult and hence we need to be careful with which variants we assess clinically and how we do this systematically and reproducibly," he said.
He added that the new framework will provide a standardized reference point for laboratories and researchers across the world to classify variants in noncoding regions in the context of rare disease.