Oxford Nanopore Technologies Ltd., the U.K. company developing portable gene sequencers, will begin providing its MinION handheld device to some customers to test, a sign it’s taking steps toward selling the instrument.
Oxford Nanopore's disposable DNA sequencer is about the size of a USB memory stick that can be plugged directly into a laptop or desktop computer and used to perform a single-molecule sensing experiment. The device is expected to sell for around $1,000, according to the company.
Oxford Nanopore Technologies also unveiled a larger benchtop version of the technology. It says a configuration of 20 of the benchtop instruments could completely sequence a human genome in 15 minutes.
The technology is based on a radically different sequencing method that has been in the work for more than a decade at Oxford University, Harvard and the University of California, Santa Cruz. DNA strands are pulled through nanopores embedded in a polymer. As the DNA passes through the nanopore, specific sequences are identified based on varying electronic signals from the different bases. As a result, the technology can read DNA sequences directly and continuously. The company says double-stranded DNA can be sensed directly from blood.
These type of nanopore machines should be suitable to sit on the bench of a small lab, running small projects and with small budgets and floorspace. However, this isn’t the full story. Each individual machine is rocking the VCR-machine-circa-1992 look, and the reason for this becomes clear when you see many of them together. The boxes are designed to fit together in standard computing cluster racks, and Oxford Nanopore refer to each of the individual machines as “nodes”. The nodes connect together via a standard network, and can talk to each other, as well as reporting data in real time through the network to other computers. When joined together like this, one machine can be designated as the control node, and during sequencing many nodes can be assigned to sequence the same sample.
Another aspect is the ability of the machines to react in real time. The sequencer can change aspects of its behavior in response to orders given during sequencing. Some of these will be automatic quality-control changes; the salt concentration and the temperature can change to optimize the sequence speed or quality. The machines can also be given basic preset targets; sequence until we have enough reads, or enough coverage, or a good enough idea of the concentration of a particular protein. This means that instead of running the machine for a set period of time, you can instead run until you have what you want.
Also, the machines can be loaded with up to 96 different samples, so you can decide to sequence one sample until you have enough DNA from it, then move onto another one, and so on. The machines can also talk to each other; for instance, four machines could sequence the same sample, and stop once they had produced enough sequence between them. Finally, the machines has built in APIs to allow them to respond to external programs of arbitrary complexity; for instance, you could connect your machines to a computing cluster that is aligning reads and making variant calls as the sequence runs, and you could decide which sample to sequence next based on the SNP calls from the first.
This new generation of sequencing machines is going to raise a whole new set of bioinformatics challanges, as well as requiring scientists to think about experimental design more carefully to make the most of this technology.