By Julia Karow
This article has been updated with additional information about optical sequencing technologies.
Researchers at the University of Washington have developed a method to generate both short reads and long-range information about their position on a DNA molecule using a standard Illumina flow cell.
The method combines the benefits of high-throughput sequencing and optical mapping in a single platform, according to Jerrod Schwartz, a postdoctoral fellow in Jay Shendure's lab at the UW department of genome sciences, who has been developing the technique for the past year or so.
Schwartz presented an outline of the optical sequencing approach last month at the Advances in Genome Biology and Technology meeting in Marco Island, Fla. A paper providing more details is currently under review at a scientific journal.
One of the limitations of short-read sequencing technologies today is that they provide no long-range contiguity, or positional information about a read on a DNA molecule. This has limited their use for the high-quality and low-cost de novo assembly of mammalian genomes, according to Schwartz.
His new method creates sequencing libraries from long DNA molecules directly on an Illumina flow cell, providing long-range information based on the distance between DNA clusters generated near the ends of the molecules.
He starts by adding flow cell adaptors to the ends of DNA fragments, currently up to about 8 kilobases long. The fragments are then added to a standard Illumina flow cell, where their ends hybridize to the surface primers.
A transposase that is loaded with the second flow cell adaptor is then added. It fragments the hybridized DNA and ligates the adaptors onto the newly created ends. This generates two smaller DNA fragments in close proximity to each other that have both flow cell adaptors attached to them.
After generating clusters through Illumina's bridge amplification protocol, both DNA fragments can be sequenced. Clusters that are close to each other on the flow cell are likely to derive from the same DNA molecule, so the relative position of the reads can be inferred.
In a variation of this method, Schwartz and his colleagues have stretched out long DNA molecules on the flow cell using an electric field. The distance between their ends, where reads are generated, relates to the length of the sequence between the reads.
Ultimately, Schwartz said, the researchers want to create multiple clusters and reads along the backbone of the same DNA molecule.
They are currently starting to apply their method to longer DNA molecules, ranging in size from 10 to 40 kilobases, and it is not clear yet what the size limit will be, he said.
In proof-of-principle experiments, the researchers have shown that the method, "works pretty well," he said, adding that high-quality de novo genome assembly, haplotyping, and sequencing through "any kind of challenging region" are among possible applications.
Illumina is "definitely aware" of the method but Schwartz said he could not comment on whether the company is interested in developing it commercially.
The concept of optical sequencing, meanwhile, is not new. David Schwartz, a professor of chemistry and genetics at the University of Wisconsin-Madison, first published an article on optical sequencing of single molecules in 2004.
Five years ago, he also published a paper on a method for mapping barcoded DNA molecules in nanoslits, which he said at the time could be developed further into optical sequencing (IS 3/13/2007). Schwartz, who is unrelated to Jerrod Schwartz, told In Sequence this week that he tried to persuade Helicos BioSciences to develop optical mapping for its sequencing platform back in 2006.
Others have developed alternative approaches to obtain both sequence and long-range positional information as well, but usually these involve more than one platform.
For example, researchers have started to combine long, low-accuracy reads from Pacific Biosciences' platform with more accurate short reads from Illumina, Ion Torrent, or 454 to improve the quality of de novo genome assemblies (IS 3/6/2012).
Optical mapping, which David Schwartz also invented, is another way to generate long-range information. OpGen, which has commercialized the technology, recently introduced a human chromosome mapping service to help detect large structural variations that next-generation sequencing alone would miss (IS 2/28/2012).
Finally, nanopore sequencing technologies promise to provide sequence information for long DNA molecules, but it is not clear yet what the quality and cost will be.
Have topics you'd like to see covered in In Sequence? Contact the editor at jkarow [at] genomeweb [.] com.