CAMBRIDGE, Mass.—GnuBIO, which seeks to be a pioneer in scalable, solution-based, low-cost DNA sequencing, announced recently that it has obtained an exclusive worldwide license for the use of microfluidics and emulsion-based methodologies for nucleic acid analysis—a novel approach developed in one of Harvard University's labs.
With the rights to this technology, GnuBIO also asserts that it will now be able to produce a DNA sequencer that has the capability to sequence the human genome at 30X coverage for a cost of less than $100, as well as the capability to run diagnostic-scale batches (tens of samples) across a small candidate gene region at 100X coverage for less than $2 per sample.
In addition to licensing technology from Harvard, GnuBIO is a Harvard-based startup company itself, formed around technology developed by David Weitz, Mallinckrodt Professor of Physics and Applied Physics in the Harvard School of Engineering and Applied Sciences and Department of Physics.
The limitations of current methodologies and technologies that sequence the human genome are that they use a two-dimensional matrix, such as a chip or bead, Weitz says, adding, "therefore, large sample numbers, and broad genomic regions, must be analyzed in order to amortize the cost of the run."
The GnuBIO microfluidics approach, on the other hand, uses microdroplets (emulsions) as microscopic test tubes, which flow at rates of millions per second, past a picoinjector that essentially serves as a pipettor and injects samples into the droplets.
John Boyce, a co-founder of GnuBIO, says this generates millions of DNA sequencing reactions per second and does not need to amortize the cost of a fixed surface such as a chip or bead. He adds that the system is not limited to DNA sequencing, as the combination of GnuBIO's picoinjector with its novel microfluidic approach will allow any assay that works in a test tube to be run within the emulsions of the GnuBIO system—at millions of reactions per second, and at a throughput orders of magnitude faster and cheaper than with current methodologies.
"Thus, the system will enable clinicians and diagnostic labs to cost effectively analyze a genomic region from a sample in minutes, compared to the current several-day turnaround time," Boyce says.
The company is setting its mark at 30X coverage for human genomes, calling that "the accepted coverage" level, though some human sequencing market watchers have noted that sequencing each base in the genome 30 times may be on the low side even for "recreational genomics," as Daniel MacArthur noted at the Genetic Future section of ScienceBlogs last year in comparing the costs of various second-generation sequencers for human genome sequencing. As he noted then, "for clinical applications, you might be looking at coverage higher than 100X."
But GnuBIO finds 30X sufficient for now for the human genome, Boyce says, noting, "Conversely, tumor samples are heterogeneous and one needs to sequence tumor DNA at a high enough coverage in order to pick up low allele frequencies—therefore, 100X coverage is a standard metric when considering sequencing tumor DNA."
He says the "real power" of the GnuBIO platform is the fact that the cost scales as a function of sample number per genomic region.
"For example, if one wanted to sequence a 1.5KB region across 12 samples at 100X coverage (assuming tumor DNA), the cost would be approximately 15 cents per sample. The cost per sample is 15 cents, regardless of whether 1 sample is run or 1,000 samples are run," Boyce says. "On another axis, if one wanted to run those same 12 samples at 100X coverage across a 750bp region (half of the size of the region above), then the sample cost would equate to $.0075 per sample."
This scalability will remove the barriers associated with DNA sequencing costs within the diagnostic and applied markets, he says.
"High-throughput DNA sequencing systems currently do not scale with sample number, and most diagnostic labs batch their samples in such numbers that make the per-sample cost too high for DNA sequencing runs," Boyce notes. "The company is rapidly working toward the production of its early access unit and will have systems available early next year. The technology has been proven, and is currently working today in the labs, therefore no invention or miracle of science is needed—it is now just a matter of scaling it up."
That confidence certainly garnered a lot of media attention for GnuBIO in June, but also raised many eyebrows in the pharma and biotech community with all the talk of $30 to $100 human genomes—particularly coming from a company that is still seeking venture capital.
In June, Boyce told media that GnuBIO was considering term sheets from venture-capital firms and would likely announce a Series A round, though he declined to say how large he expected the raise to be, or how many firms might be part of it. Reportedly, the company has sold the first two prototypes of its gene sequencer, apparently priced at $45,000, which is significantly less than others commercially available sequencers right now. It had also said at the time that it hoped to deliver beta versions of its system by late this year.
At Robodustrial.com, James Melzer expressed skepticism that such lofty goals as GnuBIO's were achievable so early in their existence and with limited funding, but added, "It beats me how they're going to do it, but like I said, leave it to the brains at Harvard to figure something like that out."
Keith Robison at the Omics! Omics! blog was a bit more verbose in his skepticism in a June 7 post, "What's Gnu in Sequencing?" in which he wrote, "It's audacious, and if they really can change out reactions within the picoliter droplets, technically it is quite a feat." But he noted that the under-$100 claim seems unreasonable because even with reducing the reagent cost so much, library construction costs would likely be somewhere in the $200 to $500 range.
"Now, perhaps they have a library prep trick up their sleeve or perhaps they can somehow go with a Helicos-style, 'look Ma, no library construction' scheme," Robison wrote. "Since they have apparently not settled on a chemistry (which will also almost certainly impose technology licensing costs—or developing a brand new chemistry—or getting the Polonator chemistry, which is touted as license-free), anything is possible—but I'd generally bet this will be a clonal sequencing scheme requiring in-droplet PCR."
