NEW YORK — Researchers at molecular diagnostics firm Color Health have uncovered hundreds of structural variants affecting genes linked to hereditary cancers or cardiovascular disorders that may be missed by sequencing approaches that do not incorporate intronic information.
The Color team used a targeted next-generation sequencing panel that included intronic regions to analyze structural variants (SVs) in genes associated with hereditary cancers and cardiovascular disorders. As they reported on Thursday in the Journal of Molecular Diagnostics, the researchers detected more than 800 structural variants affecting about 50 genes that could contribute to disease susceptibility. They also found that 80 percent of the SVs could have been missed had a whole-exome sequencing approach alone been used, which could have ramifications for some patients who have undergone hereditary cancer or cardiovascular disease testing.
"The relevance of large copy number variants to hereditary disorders has been long recognized, and population sequencing efforts have chronicled many common structural variants," first author Jeroen Van Den Akker, head of bioinformatics at Color, said in an email. "However, limited data are available on the clinical contribution of rare germline structural variants."
For their retrospective study, the researchers analyzed data from individuals who underwent physician-ordered testing for hereditary cancer or hereditary cardiovascular disease. The targeted next-generation sequencing panels encompassed 30 genes associated with hereditary cancer and 30 with hereditary cardiovascular disorders. Finding SVs in disease-linked genes like these, Van Den Akker noted, could help improve disease risk prediction.
The number of individuals tested per gene ranged between about 40,000 and 215,000. In all, the researchers identified 19,821 pathogenic variants for hereditary cancer or hereditary cardiovascular disease. This set included 1,817 structural variants, of which 1,328 were determined to be pathogenic. The researchers used a bioinformatics pipeline they developed to detect structural variants from sequencing data that included flanking intronic data and breakpoints, factors they determined to be key for robust SV calling.
While the number of structural variants was small compared to the number of SNVs, they represented an outsized share of pathogenic variants. This, the researchers noted, underscores the relevance of SVs as risk factors in hereditary cancer or hereditary cardiovascular disease.
Overall, most of the SVs — 63.8 percent — were deletions, and nearly 93 percent of the copy-number gains were simple duplications. A small percentage of SVs stemmed from mobile element insertions, though those alterations were more prominent within the BRCA2 gene, and nearly 5 percent of pathogenic variants were due to inversions or complex rearrangements.
The researchers further simulated whether the structural variants they detected would also be picked up by other platforms used in clinical sequencing, based on read depth alone. They estimated that genome sequencing at 30X coverage would be able to detect about 71 percent of SVs based on read depth, while panel testing with 500X coverage would be able to capture 53 percent and exome sequencing at 100X coverage would identify 20 percent of SVs.
This finding suggests that some testing approaches could miss structural variants that contribute to disease risk. "For people meeting professional guidelines for hereditary conditions based on personal and/or family history, it will be critical to pursue genetic testing that includes highly sensitive detection of structural variants," Van Den Akker said.
He noted that his team's novel approach for genotyping structural variants would also be helpful for large-scale population testing using whole genome sequencing. "We also expect this algorithm to perform well in specimens of lower quality, after biobanking, for example, and are actively looking for partners interested to test this," he added.