By Julia Karow
GnuBio plans to release a beta version of its desktop microfluidic sequencer in the second quarter and to launch the system commercially by the end of the year, the company said earlier this month.
The firm plans to focus on gene panel resequencing and replacing Sanger-based single-gene assays with its instrument, which integrates target amplification, sequencing, and data analysis. GnuBio promises its customers low cost, high accuracy, and fast turnaround times.
Recently, the company was also able to increase the number of barcodes it can use for its sequencing probes, allowing it in principle to obtain 64-kilobase reads, according to John Healy, GnuBio's vice president of informatics, who spoke at the Advances in Genome Biology and Technology meeting in Marco Island, Fla.
The $50,000 instrument, a beta version of which is currently under construction, will measure about 40 cm x 50 cm x 30 cm, with a weight of about 30 to 40 pounds. The fluidic system and all reagents will be contained in a microfluidic cartridge that the user inserts for each run, and no additional computing equipment will be required for base and variant calling.
The cartridge is pre-loaded with emulsions of a custom-made PCR primer library and a sequencing probe library and will cost about $200 for a 500-gene panel. Users only need to inject genomic DNA, typically 200 nanograms for a sequencing panel. A run takes about three and a half hours for a 500-gene panel at 1,000x coverage, including on-board variant calling, and generates about 1.2 gigabases of data.
The system first amplifies the target DNA in single droplet-based PCR reactions and then reads the sequence of each amplicon by measuring the hybridization of fluorescently labeled random hexamer probes. Each probe comes in a droplet with a color that distinguishes it from all the other probes in the library. Only probes that match exactly act as primers for a DNA polymerase that displaces a quencher probe, enabling the fluorescent signal to be read (IS 10/4/2011). At the heart of the system is picoinjector technology that GnuBio licenses exclusively from Harvard University, where it was developed in David Weitz's lab.
Base and variant calling happen on the instrument in real time. After generating overlap reads from the hybridized hexamer probes and mapping those to the genome, a "very loud signal" in the area of the target DNA emerges, Healy said. The next step is to map the matching probes to the target area, and to use probes that did not bind at a specific position as "negative evidence" to help identify true variants if there is ambiguity.
With a hexamer probe library — which actually also contains a few heptamers, octamers, and nonamers in order to cover homopolymers and dinucleotide repeat sequences — the theoretical read length limit is about 1 kilobase, Healy explained, though the company has so far only tried sequencing amplicons up to 612 base pairs in length.
So far, company researchers and their collaborators have achieved reads with a per-base accuracy of at least Q70, and have been able to read six-base homopolymers and nine-base dinucleotide repeats. Sequencing errors are independent of the base position in the read.
About a month ago, Healy reported, the company also managed to increase the number of colors, or "dye barcodes," that identify each hybridization probe from 15,000 to at least 300,000.
Up until now, each barcode color resulted from a mix of four different dyes, but now, the company can add two more dyes, each at five different levels, for a total of 300,000 different barcodes. With those, it can go from hexamer to nonamer probes in the future, increasing the upper read length from 1 kilobase to 64 kilobases. This will enable the company to interrogate haplotypes and large structural variations.
As a result, "whole-genome applications, as we scale up, are a much closer target than we originally thought," Healy said. Prior to the increase in dyes, "we thought this problem was going to be much harder than it is."
In terms of applications for its instrument, GnuBio plans to focus on resequencing gene panels — competing with systems like Ion Torrent's PGM — and on replacing Sanger sequencing of single genes.
Because its system integrates target amplification and sequencing, with no need to break emulsions, its target coverage is even and does not suffer from library prep and system bias, the company claims.
A resequencing assay could be ready within about a week and a half, according to Healy, starting with the customer ordering an assay for specific targets, seven days of PCR primer production by an outside oligo provider, two days of library production and cartridge pre-loading at GnuBio, and two days of validation if requested. After starting with 200-gene panels this year, the company plans to launch a 1,000-gene cartridge in early 2013.
For single-gene sequencing, currently planned for early next year, customers would inject a sample with their own primers into a reusable cartridge and retrieve their sequence within an hour and a half at a cost of about 50 cents per sample, according to the firm.
In the second quarter, the company plans to ship four beta systems to early-access customers, which it has not yet selected. Following that program, it plans to launch the system commercially by the end of this year.
So far, GnuBio has been collaborating with a number of labs, including the City of Hope Molecular Diagnostic Laboratory and the Université de Montréal Pharmacogenomics Centre at the Montreal Heart institute. The company is currently working on a manuscript for a peer-reviewed publication with one of its partners, which will detail not only its technology but also present a "biologically significant result," according to the firm.
Last month, GnuBio said it is working with the Montreal Heart institute, which already has an early-access system, to develop a gene panel to determine the risk for sudden cardiac death (PGx Reporter 2/1/2012).
Have topics you'd like to see covered in In Sequence? Contact the editor at jkarow [at] genomeweb [.] com.