NEW YORK (GenomeWeb) – Researchers at the Institute of Environmental Science and Research (ESR) in Upper Hutt, New Zealand, have tested the Oxford Nanopore MinIon for sequencing the influenza virus and are planning to use the device for virus detection and discovery in metagenomic samples.
Even though the MinIon influenza sequence data, generated in a single four-hour run, was more than 99 percent identical to data obtained from the Illumina MiSeq and from Sanger sequencing, clinical applications will require 100 percent concordance with conventional sequence data, the authors noted. But the portability and lack of infrastructure requirements of the MinIon make it promising for in-field surveillance applications, such as during viral outbreaks.
The ESR team, known as the "Virus Hunters", published the influenza genome study online in Frontiers in Microbiology this week. Nicole Moore, one of the authors, presented the main results at Oxford Nanopore's "London Calling" conference in May.
The group, which started in 2009, has been using metagenomics to discover new viruses and detect known viruses in human, animal, or environmental samples, for example from unsolved gastroenteritis outbreaks, native bats, and aerosols from animal slaughterhouses. In addition, it has been sequencing viral genomes, initially with the Roche 454 GS Junior but lately with the Illumina MiSeq.
Last year, the researchers enrolled in Oxford Nanopore's early-access program for the MinIon. "We were interested more in the portability of the instrument than its ability to do long reads, particularly for metagenomic virus discovery," Richard Hall, a Virus Hunters team member and the senior author of the paper, told GenomeWeb.
For their initial project, they sequenced the genome of an influenza A isolate on the MinIon, the Illumina MiSeq, and by Sanger sequencing. The sample came from New Zealand's World Health Organization National Influenza Centre, which is based at ESR. For all three sequencing platforms, the influenza RNA genome, which has a total length of 11 kilobases and occurs in eight separate gene segments, was amplified using a one-tube RT-PCR protocol, and sequencing libraries were prepared from the amplicons.
The MinIon run, which used flow cell version R7.3, took four hours and generated about 118,000 sequence reads, of which 17,100, or about 15 percent, were high-quality 2D reads, with a mean read length of 955 base pairs. Almost 90 percent of these high-quality reads aligned to the influenza genome, covering the genome completely. Including library preparation, the MinIon workflow took about eight hours, a "clinically relevant timeframe," Hall said.
The MiSeq run delivered about 3 million reads of 250 base pair length, which also covered the influenza genome completely. While the sequence depth was much higher for the MiSeq than the MinION due to the larger number of reads, the MinIon data covered most of the gene segments more evenly than the MiSeq.
Overall, the MinIon influenza sequence was more than 99 percent identical to both the MiSeq and the Sanger data, while the MiSeq and Sanger sequences agreed 100 percent. Most likely, the differences, which were distributed randomly across the genome, resulted from sequencing errors of the MinIon. "Getting rid of those last few discrepancies is really important, particularly for clinical applications," Hall said, because even small differences could be important for things like vaccine response and resistance to antiviral drugs.
Just running the MinIon for longer to increase sequence depth might improve the consensus accuracy, he said. In addition, the researchers are looking forward to receiving their first flow cells for the MinIon MkI, which Oxford Nanopore recently launched and which promises improved performance.
Hall also said he is interested in the automated sample prep system, called Voltrax, that Oxford Nanopore announced in May, as well as in improvements to the analysis software package.
Going forward, his team plans to use the MinIon for virus discovery and unbiased virus detection in metagenomic samples and to explore its portability in the field. "With the increases in quality we've seen during the [MinIon] access program, it is now possible to take a read from a MinIon dataset, use conventional BLAST analysis, and if it's a known virus, there will be a match in the database," he said.
In the future, the MinIon could be helpful in the local surveillance of viruses, such as human and animal influenza strains, so samples would no longer needed be sent to centralized laboratories, he said. Other groups have already used the MinIon to track a Salmonella outbreak in a UK hospital, to identify viruses from metagenomic samples, to study the transmission of Ebola virus in West Africa, and to generate microbial drug-resistance profiles.
Whether or not the device might find its way into point-of-care testing for patients is still an open question, though, as sequence data may not always be needed. "I think there is still a place for PCR because often we just want to know 'Does the patient have it or not?'" he said. "Sometimes, the associated [sequence] data might be more of public health interest. But sometimes, it might be very relevant to the patient, for example in terms of drug resistance."