At the Consumer Genetics Conference last week in Boston, GnuBio unveiled its $50,000 desktop sequencer, which recently shipped to the first of the company's beta testers ahead of a wider commercial launch of the system expected next year.
GnuBio CEO John Boyce announced at the opening of the conference that a planned live demonstration of the instrument had been cancelled due to issues with conference site regulators concerning the use of live biologic materials.
Instead of a live demo, the company offered a few updates on the system and discussed the machine's capabilities in light of feedback from the clinical lab market. While GnuBio has said previously that the price for sequencing a sample using a 50-gene panel cartridge would be near $50, Boyce cited an "all-in" price of $200 per 50-gene cartridge in his CGC presentation.
He also reported accuracy data the company achieved on a set of 24 runs from eight germline DNA samples using a panel of 36 genes associated with the Ashkenazi Jewish population. In this test, the sequence error rate based on unfiltered raw data was about 0.0005, Boyce said. Filtered to remove no-calls — which represent about 4 percent of the total — the error rate went down to zero, Boyce said.
GnuBio's system is based on "picoinjector" technology developed in co-founder David Weitz's physics laboratory at Harvard University (IS 10/4/2011).
The company describes the instrument as sample-to-answer, although some sample prep is needed to isolate genomic DNA for input into the machine's microfluidic cartridges. After DNA is pipetted into a cartridge it is sheared into 1-kilobase sized fragments, which a picoinjector injects into microdroplets that contain PCR primers representing the amplicon to be sequenced.
After the DNA is enriched, the droplets meet up with a second stream containing fluorescently labeled hexamers. Only the hexamers that match exactly are able to hybridize to the DNA and act as primers for DNA polymerase, which turns on the fluorescence and enables the sequence to be read.
Boyce said the picoinjector allows 10,000 uniplex reactions per second, which in turn allow the machine to use 1/1000 the reagents required of other sequencers.
The hexamer system limits the read length, currently, to 1-kbp, the company said, but this could be extended by expanding the set of oligos it uses to construct its library with longer molecules.
While Boyce stressed that plans for a wide launch of the system are still up in the air, he and other GnuBio representatives answering questions at the conference said that the company hopes to release the sequencer by summer or fall of 2014.
He also said the company has grown to 62 employees from 28 in late 2012.
GnuBio is still planning to offer targeted sequencing panels for the system of up to 50 genes with the ability to detect alleles at frequencies down to five percent, which Boyce said would cost $200 per cartridge "all-in" for either pre-developed panels, slightly modified versions of pre-developed panels, or custom panels, all of which will run on the company's instrument in 3.5 hours or less.
"We have ways to bring that cost down in the future," he added, "but that's the start."
Boyce did not detail what pre-developed panels the company might release either ahead of or concurrent with the wider launch of its sequencer, apart from the Ashkenazi Jewish panel he mentioned in discussion of the system's error rate.
Boyce said that feedback from clinical laboratories has indicated that many potential customers will be less interested in targeted gene panels than in hotspot sequencing— another application of the technology that GnuBio is promoting.
"I used to talk a lot about gene panels, but I still got push back that it's still too much information [for clinical use]," Boyce said in his talk. "Clinicians want only actionable data, otherwise it’s a liability."
Based on this feedback, he said the company is planning to focus rollout of the system more heavily on this hotspot sequencing capability. He mentioned specifically in his presentation an oncology-focused hotspot panel based on the current slate of FDA-approved drugs and the mutations that those drugs target.
As the number of currently approved targeted drugs is far outweighed by the pipeline of drugs now awaiting or approaching approval, Boyce said the company will add to this panel as time goes on, with the potential to target mutations down to one percent allele frequency.
While GnuBio previously said that it planned to ship four beta systems to customers in 2012 (IS 2/28/2012), its first beta system only recently went out as of April 2013, Boyce said. He declined to provide details about this initial beta tester, or whether and when the company may add additional labs to its roster of early testers.
GnuBio's Executive Vice President of Product Development Sepehr Kiani, who answered questions about the instrument at the conference, told attendees that the current beta tester is a CLIA lab, and that they are using the machine for hotspot interrogation rather than looking at gene panels.
The company has said it is looking to eventually gain US Food and Drug Administration 510(k) clearance for its system, likely by partnering with a diagnostic company on specific assay design. In the meantime, when the product is launched — potentially next year — it will be for research use only, but also targeting labs interested in using it for laboratory-developed-tests in the CLIA environment.