BioNano Genomics has launched its nanochannel-based single-molecule DNA mapping platform in Europe and plans to extend the technology to human-size genomes later this year, enabling researchers to discover new structural variants, separate haplotypes, and obtain better genome assemblies.
At the European Human Genetics Conference in Paris this week, Pui-Yan Kwok, a professor at the University of California, San Francisco, School of Medicine and an early BioNano customer and collaborator, presented the first two human genome maps constructed with the BioNano technology.
BioNano launched its Irys DNA mapping platform last November, initially in North America and China, and primarily for mid-size genomes up to about 300 megabases in size. In a proof-of-concept study last year, the company and its collaborators showed that they could use the technology to generate haplotype-resolved sequence motif maps for the human MHC region, but experts criticized at the time that the researchers were unable to analyze human genomic DNA (IS 7/17/2012).
The Irys system currently uses chips with three sets of several thousand nanochannel arrays. Individual DNA molecules up to about 500 kilobases in length that are fluorescently labeled at specific sequence motifs by a nicking enzyme are drawn into the channels by a voltage, held in their stretched-out position, imaged automatically, and released. Because the DNA molecules are stretched uniformly in the channels, the distance between the labels can be measured, and sequence motif maps can be constructed from the data, similar to restriction enzyme maps. These maps can then serve as scaffolds for the assembly of short-read sequence data.
Because the DNA molecules are so long, and the technology allows for several labels with different colors to be used, "repeats and complicated regions are actually spanned by one molecule, making it possible to cover even very large structural variants," Kwok told In Sequence.
He and his colleagues used the Irys system to generate sequence motif maps at 50x coverage for two human genomes that have been extensively characterized as part of the HapMap and 1000 Genomes Project: the father and daughter of the CEPH-CEU trio, samples NA12891 and NA12878.
They generated the maps from DNA labeled at the Nt.BspQI enzyme sites that ranged in length from 150 kilobases to 500 kilobases. They have been using the maps to resolve haplotypes, for example in the MHC region; identify known and novel structural variants; and to help with the de novo sequence assembly of the genomes.
Kwok said the researchers have identified "quite a few molecules" that do not map to the human reference genome. Because those carry repeat patterns, they are probably parts of centromeres and telomeres, "regions not accessible to short-read sequencing methods," he said. In addition, they have found large duplications and inversions "that are very hard to figure out in the genome by other methods."
He and his colleagues are hoping to publish their results later this year. They also plan to use the Irys system to map the samples of the 1000 Genomes Project in order to provide a set of reference maps for their de novo assembly. In addition, they want to map the genomes of smaller organisms that have already been sequenced but could not be assembled because of problems with repeats and short reads.
Todd Dickinson, BioNano's vice president of global commercial operations, said the human data presented by Kwok was collected on the existing chips, which would be impractical for customers because so many chips are needed for a human genome. The company is working on a new version of the chip with more channels that will have eight to 10 times the throughput of the current chip, he said, which will be "a key component" of its ability to support human genome analysis later this year.
The company has also been making improvements to its platform and assay since the beta testing phase, including a better autofocus, being able to use longer DNA molecules of higher quality, and improved software for image analysis and map assembly.
Beta testing of the platform involved three groups: Kwok's lab at UCSF, a group at Kansas State University, and Genergy Biotechnology, a genomic services company in Shanghai that is a sister company of BioNano's China distributor. All three purchased their Irys systems following the completion of the beta program, and BioNano is now working on expanding its customer base, targeting both genome centers and smaller users.
A few weeks ago, it placed its first instrument with a currently undisclosed customer in Europe, Dickinson said. To support its launch in Europe, which is happening earlier than planned because of strong interest from researchers, it also hired a vice president of commercial operations for Europe, the Middle East, and Africa (IS 6/4/2013).
BioNano, which raised $10 million in a Series B-1 financing round a year ago and is considering raising additional funding in the near future, currently sees structural variation analysis as the primary application for its technology, followed by genome assembly and finishing. Haplotype phasing will also likely be an important application, Dickinson said, and the firm is currently working on phasing an entire human genome. Ultimately, the technology could also be used for epigenomics analyses, he said.
"We're really excited about the potential for this platform in disease research, in particular, cancer," Dickinson said, where structural variants are known to play an important role.
Apart from the human genome market, where BioNano sees the greatest opportunity, its platform might also be useful for assembling genomes that do not have a reference, and for analyzing complex regions within plant genomes. Earlier this year, for example, company researchers, in collaboration with a group at the US Department of Agriculture and others, published a paper in PLoS where they used sequence motif maps to help assemble a highly repetitive region in the genome of a wheat-related grass.
But the company faces competition from others that are targeting the long-range genome structure market, among them OpGen and Nabsys.
Dickinson acknowledged that OpGen "is probably the technology that paved the way for this genome mapping space," but he said that BioNano's nanochannels allow it to stretch out DNA "in a much more uniform and a much more high-throughput way" than OpGen.
And while he said Nabsys has "a really cool technology," it is not fully commercialized yet and will focus on smaller genomes initially (IS 1/15/2013).
He said the company does not regard Illumina's Moleculo technology as competition because that only extends the read length to up about 10 kilobases, much shorter than the molecules BioNano maps. Rather, Moleculo reads could complement its own data, he said.