This article has been updated to include comments from an early access customer.
MARCO ISLAND, Florida (GenomeWeb) – Genapsys plans to launch its portable next-generation sequencing instrument in the second half of this year, Tyson Clark, Genapsys' director of applications, said in a presentation at the Advances in Genome Biology and Technology meeting here on Thursday.
In addition, Clark discussed data that has been generated internally and with early-access customers on sample types including cell lines, microbes, formalin-fixed paraffin-embedded tissue, and cell-free DNA.
As previously reported at the JP Morgan Healthcare conference earlier this year, the 10-inch by 7-inch system, dubbed Genius, will cost around $10,000 and will include three chips: one with 1 million sensors, which will cost $100 per run; a second with 16 million sensors, which will cost $300 per run; and a third with 144 million sensors that will cost $600 per run.
The technology detects base incorporation using an electronic signal. Genapsys CEO Hesaam Esfandyarpour said in an interview at AGBT that clonally amplified DNA molecules sit within a detection chamber containing CMOS sensors. A current is continually passed through the space and nucleotides are injected into the space one at a time. If the nucleotide binds to the DNA molecule, the electronic sensor is able to detect the incorporation, Esfandyarpour said.
Samples are prepared in the same way as other NGS technologies, but with a clonal amplification step at the end. The prepared library is loaded into the Genius instrument. Esfandyarpour declined to disclose additional details about the sequencing and sample prep methods, but noted that forthcoming papers would go into more detail.
Read lengths at launch will be between 100 bases and 150 bases and will be single-end sequencing, but Esfandyarpour noted that in the future, read lengths would increase. Clark added that the company would also enable paired-end sequencing in the future.
During Clark's presentation, he said that the company sequenced an Escherichia coli sample using its medium-throughput chip. Raw read accuracy was greater than 99 percent and the chip had between 12 million and 15 million active sensors. Sequencing demonstrated even coverage across the genome.
As previously reported, Genapsys also sequenced the well-characterized NA12878 sample and calculated error rates for substitutions, deletions, and insertions at 0.01 percent, 0.03 percent, and 0.04 percent, respectively, when doing exome sequencing using a probe-based capture kit from Integrated DNA Technologies. They also compared the data with Illumina data, finding that they were 93 percent concordant.
Clark also noted that the system could identify somatic mutations at allele frequencies down to 1 percent and demonstrated that the system could work with both IDT's hybridization capture panel as well as Thermo Fisher Scientific's AmpliSeq hotspot panel.
Clark also noted that the company has tested the system on cell-free DNA from two transplant patients and two controls. The system was able to distinguish the normal controls from the transplant patients and also distinguish between a patient undergoing rejection and one who was not, with good correlation with Illumina sequencing data.
In a collaboration with Chinese service provider Novogene, researchers have tested the Genapsys system for sequencing the exomes of cancer patient FFPE samples. With the Genapsys system, 99.5 percent of the data was considered Q20 or higher quality and 84.9 percent was at Q30.
When researchers compared Novogene's standard exome sequencing pipeline, which uses Agilent SureSelect technology and Illumina sequencing with the IDT exome kit and Genapsys sequencing, they found that the two were highly correlated in terms of the variants called.
Other early-access users include Stanford Medicine's Hanlee Ji, who has done more than 10 runs on the system, Clark said, to sequence circulating tumor DNA.
In addition, the Jackson Laboratory's Chia-Lin Wei has evaluated the technology on a 212-gene targeted cancer panel, and the HudsonAlpha Institute of Biotechnology's Brian Roberts has been running a number of different samples on the system, and achieving error rates between 0.3 percent and 0.7 percent, Clark said.
Wei said on the sidelines of the AGBT meeting that her group began testing the instrument about three weeks ago and has two running in the lab. She said that while the team has not yet completed doing a thorough analysis of the data, the "quality seems pretty good, especially considering that it is an early platform."
She added that her lab plans to work with Genapsys to perform benchmark calibrations of output, quality, error rate, and biases and will publish that data.
Currently, she is running the mid-throughput chip and achieving between 10 million and 12 million single-end reads. That chip is suitable for targeted panels, and Wei is particularly interested in using it for cancer panels. Ultimately, if it performs well, the system's "small footprint and reasonable cost could make it useful for clinical sequencing mutation detection," she said.
Wei anticipated that the current market for the Genius system would be targeted panels, microbial genomes, and cell-free DNA — applications where it is not critical to have paired-end reads.
In the future, Esfandyarpour said the firm plans to launch an automated sample-prep instrument that would be compatible with most commercially available library prep technologies. The company also plans to develop additional sensing capabilities so the instrument can be used to detect methylation, proteins, and live cells. He did not provide a timeline for these additional developments.