This story has been updated from a previous version to correctly report that the Irys technology was used to detect large heterozygous variants, not single nucleotide variants.
NEW YORK (GenomeWeb) – BioNano Genomics in recent years has been working with collaborators to demonstrate that its Irys System for physical genome mapping is an important tool for human genomic research.
At the American Society for Human Genetics annual meeting in October, BioNano and partners at various research institutions presented several posters showing that Irys can be a very useful tool in variant research, independent of sequencing technologies, BioNano CEO Erik Holmlin told GenomeWeb in an interview following the meeting.
BioNano's Irys System uses what the company refers to as next-generation mapping (NGM). The system is a nanochannel-based single-molecule DNA mapping platform that uses chips that contain several thousand nanochannel arrays. It measures the distance between fluorescently labeled sections of DNA and uses that data to construct de novo map assemblies that can serve as scaffolds for NGS data. "We are primarily selling to researchers," Holmlin said. "It's really a translational research tool." However, he said that exploring uses for it in the clinical market "is very much in our future."
The company fills a unique niche in the genomics market, since what it offers isn't a sequencing technology but a physical mapping technology. It's not new, but other companies who have tried to exist in the field, including OpGen and Nabsys, weren't able to scale their technology to look at large complex genomes, such as the human genome, Holmlin said. BioNano owns intellectual property in nanotechnology that allowed the company to rapidly scale up its technology and work on human genomes, he added.
"Physical mapping is a complement to sequencing," Holmlin explained. But the company also wants to be "on par" with existing technologies and is working hard to match the high-throughput of Illumina's systems, he said.
To date, the company has primarily marketed Irys as a tool that researchers can use alongside next-generation sequencing, particularly when they are building reference genomes. BioNano has worked with researchers interested in building reference genomes for hummingbirds and the duck-billed platypus, but has also worked with researchers to increase reference genome diversity in humans, Holmlin said.
"In humans we have reference genomes, but they don't [represent human] diversity," Holmlin explained. To that end, the company helped researchers use Irys to create a de novo assembly of a Korean individual (AK1), work that was published earlier this month in Nature.
But the presentations at ASHG show that it can be used independently of sequencing, and in fact meets what Holmlin describes as a "critical unmet need in human genomics," the ability to measure genome structure.
This is particularly important when researchers are studying structural variation. For example, Vanessa Hayes, a researcher in comparative and prostate cancer genomics at the Garvan Institute of Medical Research and an author of one of the ASHG posters, discovered 15 new structural variants in prostate cancer using NGM on the Irys System.
As a prostate cancer researcher looking towards precision medicine approaches to treating the disease, Hayes felt that NGS alone couldn't identify all the structural variations. "When it comes to these large complex rearrangements and structural variations, essentially we are inferring [variants] by using NGS," Hayes said. "We are not observing them."
Hayes and her team generated NGM data on the Irys System and NGS data on the Illumina HiSeq X10 and clearly showed that NGS wasn't catching all the variants. "Without NGM we missed 90 percent of the [variant calls]," she said.
This is because unlike short-read sequencing technologies, NGM is not limited by size and can look at much larger and more complex regions and identify variants there, she explained.
Eric Vilain, a professor of human genetics and pediatrics at the University of California, Los Angeles School of Medicine, found the Irys System similarly useful. He also presented research at ASHG showing that Irys was able to detect large heterozygous variants, and small insertions and deletions in patients testing for the inherited disease Duchenne muscular dystrophy (DMD). Vilain and his colleagues used NGM to identify heterozygous structural variants in carriers' mothers for DMD, and identified 800 insertions, 500 deletions, 100 inversions, and 50 translocations on average using the technology.
"BioNano seemed to be the main (and maybe only) technology mature enough to detect large heterozygous SVs with sufficient sensitivity and specificity to be useful for our purposes," Vilain told GenomeWeb in an email.
He added that while there were some "growing pains" involved in implementing BioNano's System, the company made sure these issues were resolved quickly.
Vilain has also continued to use the technology in his current research. "We are currently running a large number of families with undiagnosed genetic disease on the BioNano Irys system, as part of the Undiagnosed Diseases Network," he said. "So far we have identified a number of interesting SVs which we are now validating using other technologies."
In both Vilain and Hayes' research, NGM has provided important insight into structural variations that they both agree cannot be addressed with current sequencing technology. It's an important gap in current genomics research that has been pretty widely recognized within the community.
"Sequencing people have recognized [this gap], but haven't been able to address it," Holmlin explained. To be able to see these structural elements through sequencing, researchers would need reads that are hundreds of thousands of base pairs long, he said.
Pacific Biosciences, which collaborated with BioNano on the aforementioned Korean sequence assembly, is one company trying to uncover structural elements of genomes by developing techniques to produce long read sequences. While PacBio has been able to create longer reads, they are limited to 10,000 base pairs or less, Holmlin said.
In contrast, the Irys system can map "millions of base pairs," he said.
Hayes added that PacBio's technology is also not a cost-effective option for researchers compared to BioNano. NGM isn't an expensive technology like early NGS was, which gives Hayes and her colleagues more room to play around with different techniques, she said.
However, while NGM holds a lot of potential to fill gaps that sequencing cannot, Hayes cautions that NGM alone isn't going to be sufficient for all structural variant research needs. "Without the NGS, we can't actually tell what the bases are," she said. "NGM is telling you where to look. It doesn't tell you where to start in [with gene targets]."