By Julia Karow
NobleGen Biosciences has automated the DNA conversion process that underlies its "optipore" sequencing technology and will shortly install a second-generation prototype nanopore reader in its lab.
At the Consumer Genetics Conference in Boston last week, NobleGen CEO Frank Feist reported on the progress of the firm's sequencing technology since the startup was founded last year (IS 3/25/2010) and made some projections about its ultimate capabilities.
The technology, which combines solid-state nanopores with optical detection, and separates the sequencing chemistry from the readout, originates from the lab of Amit Meller, a professor of biomedical engineering and physics at Boston University who previously worked at Harvard. Last year, NobleGen obtained an exclusive license to the intellectual property from Harvard and BU and has been developing the technology with seed funding and grants to Meller's lab from the National Institutes of Health.
Over the last year, the firm has signed up a number of advisors that are well known in the DNA sequencing technology field, including George Church from Harvard Medical School, Chad Nusbaum from the Broad Institute, and Harold Swerdlow from the Wellcome Trust Sanger Institute (see Paired Ends, this issue).
In order to sequence DNA, NobleGen first converts the molecule into what it calls an "expanded synthetic representation," a longer molecule in which each base is replaced by a specific oligonucleotide. It then hybridizes color-coded molecular beacons to the converted DNA that each carry both a fluorophore and a quencher.
The sample is then analyzed on a benchtop nanopore sequencer that consists of a nanopore array with microfluidics, lasers, total-internal-reflection-fluorescence optics, and a charge-coupled device camera. As the DNA is pulled through a nanopore, driven by a voltage, the beacons are stripped off and the quenchers are separated from the fluorophores, generating a flash of light near the pore that is recorded by the camera. The color of the light is then translated into the corresponding base.
The sequencer has "no moving parts" and is "relatively inexpensive to put together," Feist said, and the nanopore chip at its heart is small — less than 0.5 millimeters "in the largest configuration of the system."
NobleGen has now automated the DNA conversion process on a standard liquid handling system. The four-step process includes common enzymatic reactions, such as ligases and restriction enzymes. It can be performed in bulk for an entire human genome and involves no amplification. The total conversion for a human genome takes about five hours, Feist said, and can be done for 96 samples in parallel on a microtiter plate.
So far, the company has shown that it can convert up to 16 consecutive bases, in four cycles of up to four bases each, which is equivalent to a read length of 16 bases. The goal is to increase that read length to 80 bases by the end of the year, Feist said.
In the next few weeks, NobleGen will also replace its first-generation prototype sequencer with a second-generation instrument. Feist told In Sequence that this machine will initially use an 8x8 nanopore array but will not be limited to that size. It will have custom optics that are simpler and offer higher performance than the off-the-shelf microscope of the first prototype, he said, and will read four different fluorophores instead of two.
One of the key advantages of NobleGen's platform is its high signal-to-noise ratio, he said, allowing it to work with picomolar concentrations of DNA and to read its sequence at high speeds of more than 100 bases per second per nanopore.
[ pagebreak ]
With a 400 x 400 nanopore array, the system could sequence 96 human genomes in 17 hours, or a single genome in less than 30 minutes, he said, enabled by the fact that each pore can analyze many DNA molecules sequentially.
NobleGen wants to target clinical sequencing tests with its platform, Feist said, and is confident that it will be able to achieve a turnaround time of less than 24 hours for sequencing and assembling a human genome for "significantly less" than $2,000 in total costs, including labor and IT. The firm is currently working with its advisors to develop bioinformatic approaches "that are optimally matched to assemble these data," he said.
Most clinical tests, he said, only require a read length of 200 base pairs, which the company is targeting. "It's not that we can't do more than 200, but with that, we believe we have the best competitive advantage," he said.
He also said the system will be able to produce data of sufficient quality for clinical tests, or an accuracy of Q50. The company has started to model the errors for each step of its process and has validated that model with experimental data. "From that, we believe we have a very competitive raw base error rate," he said.
Generally speaking, single-molecule techniques tend to have more deletion errors than other systems, he added, but "we have a way to deal very effectively with deletions from a bioinformatics perspective," and "the Q50 accuracy that we target seems very achievable from what we know today."
Whole-genome sequencing data, he said, could be obtained from the DNA of fewer than 1,000 cells, allowing the platform to analyze small tumor samples, FFPE samples, and DNA from cell-free samples.
Feist did not mention when NobleGen plans to commercialize its system, but has said previously that it might be able to go to market around 2014.
Have topics you'd like to see covered in In Sequence? Contact the editor at jkarow [at] genomeweb [.] com.