This article, originally published March 18, has been updated with information from Oxford Nanopore Technologies that clarifies the beginning of MinIon shipments.
Oxford Nanopore Technologies said that with its platform more than 50 kilobases of sequence can be "easily obtained from single reads," which are limited in length by the size of the input DNA rather than the sequencing system.
In a presentation at the Plant Genomics Congress in Kuala Lumpur, Malaysia, last month, Chief Technology Officer Clive Brown provided several details about the firm's nanopore technology. His presentation slides were posted on the conference website last week.
According to the presentation, the quality of the nanopore reads is independent of the read length, with "no difference between base 1 and base 50,000." Oxford Nanopore has generated reads spanning the entire bacteriophage lambda genome – 48 kilobases – on its MinIon system, according to the slides, reading 30 bases per second and nanopore.
ONT has been using bacteriophage lambda DNA as a standard control, sequencing it as part of a "routine development run performed daily on MinIons," that generates "many reads [of] 10s of [kilobases]."
At the Advances in Genome Biology and Technology conference last month, David Jaffe of the Broad Institute reported a mean read length of about 5 kilobases for two bacterial genomes sequenced using Oxford Nanopore's MinIon, with some reads up to about 15 kilobases long. But according to ONT's presentation, those read lengths reflected the size distribution of the input DNA, as shown by Agilent Bioanalyzer data.
The presentation also shed some light on the error profile of ONT's nanopore data. Overall, deletion errors are more common than insertion errors. The majority of errors can be resolved by increasing the read depth, according to the company, though there is "occasional systematic error" that the firm believes can be "resolved with [an] improved model."
As reported by Jaffe, there are "some perfect stretches" of sequence within the reads. Also, the consensus accuracy can be improved by using two different types of nanopores with different systematic error types, and according to the presentation, a "two-pore system [is a] likely product feature."
The presentation also explains how tethering the sample DNA to the membrane at one end – a technique developed and patented by ONT to concentrate the DNA near the pore – increases the sensitivity 20,000-fold and allows for the use of native DNA that is not amplified, although amplified material can be used as well.
For sequencing, the DNA is fed through an undisclosed protein nanopore by a proprietary processive "motor" enzyme. According to the presentation, this can happen in two different ways: in one mode, the DNA travels through the pore while being held back by the enzyme at one end and is then pulled back against the electric field. In the other mode, the enzyme pushes the DNA in the same direction as the field. "Both systems can traverse a hairpin and sequence both strands" of DNA, according to the company.
Brown also addressed how many of the 512 MinIon pores are active during a run. According to the presentation, "most pores [are] sequencing correctly most of the time," and data from four four-hour runs showed that half to three quarters of the pores are sequencing at the same time.
Sample preparation for the MinIon starts with either high molecular weight genomic DNA, cDNA, or amplified DNA and takes less than one hour. According to the presentation, several library prep kits are available, which both involve the addition of adapters, tethers, and enzyme "motors."
Regarding the GridIon system, a higher-throughput nanopore sequencer with more pores and longer run times that ONT plans to commercialize after the MinIon, the presentation stated that it is "in late development" and will offer "cheaper cost per base."
In the meantime, the company is getting ready to ship MinIon devices and reagents to the first participants of its early-access program, several hundred in total.
ONT originally intended to send the first sets in the second half of March, but according to a company spokesperson, shipment dates will likely be pushed back by about 10 days because more than 100 asked for an extension of the payment deadline, which ONT granted.