MinION: A complete DNA sequencer on a USB stick
A little over a year ago, the MinION USB stick DNA sequencer was announced by Oxford Nanopore Technologies. This dream device was to cost under $1,000 and be able to sequence up to 150 million base pairs in its six hour lifetime, a feature set that would allow it to leapfrog any competing technology out there. Eager followers have gotten wind that the developers have been able to work out early reliability issues and that the goal is to bring a device to market later this year. Having a product like this would give unprecedented power to patient self-quantifiers, DIY biohackers, and perhaps even DNA micromachinists — provided our medical regulatory overlords see it fit.
The MinION and its even more powerful cousin, the GridION, use a relatively new gene analysis technique known as nanopore sequencing. The actual specs reported for the release version of the MinION have undergone several revisions, but it looks like we can expect anywhere from 2,000 to 8,000 individual ion pores on-board. The pores are built into a special ASIC (application specific integrated circuit) chip made by an undisclosed defense manufacturer in San Diego. The MinION works together with your PC and is a one-shot deal. Once the device has been activated, you do your sequencing and then the device is depleted, permanently. For bigger runs the GridION system uses a stand alone machine together with individual sample cartridges, similar in size to the MinION. Additionally, it is has been reported that the GridION system will be able to analyze RNA and protein samples — a huge new area just beginning to be tapped, as Thinking Machines founder Danny Hillis described in a recent TED talk.
To read the DNA sequence, a pre-digested sample is loaded onto the chip and strands of various lengths associate with each pore. An enzyme linked to each pore separates the paired DNA strands as they progress like a ticker-tape through the pores. As they do so, ion currents, which continuously run through the pores in the background, are uniquely modulated as each base pair makes its way through. The speed of these nanomachines is astounding — each nanopore is sending about 33,000 measurements per second. The resulting signals are amplified and sent to an FPGA where things are sorted out. Some error checking is built in, since after the first half of the DNA strand is read, the complementary strand is also then pulled through in reverse order.
At the industry standard coverage of 30 reads, a complete human genome would require several of these USB sticks to run in tandem and would take a few days. Whole genomes however, are not the principal market for the MinION — large sequencing houses provide increasingly affordable sequencing and it is a safe bet that anyone who really wants their sequence will soon be able to get it. On the other hand, sequencing something like a Phi X phage, a DNA virus just 5.4 kilobases long, could be done in a single shot.
While the popular concept of a genome is that of a fixed sequence, the larger reality is that chromosomes change, both through mutation and by gross rearrangements to their structure and number. They break and fuse in both healthy and diseased cells. Monitoring these changes in different cell populations in the body might be expected to be among the central tenets of any high-tech billionaire’s plan for eternal life. The ability to analyze and separate DNA in a mixed sample of human cells together with the bacteria and viruses that invaded them would be an even greater power we will soon also demand of devices like this.
Current DNA techniques fail to capture important sequence information known in the business as phase. Phase information tells which parental chromosome a particular gene, or sequence, comes from. Provided the strings fed to each pore are long enough, nanopore sequencing could potentially also derive this information, giving further insight. Not only would heredity and genealogy benefit from having this information, but also the sometimes controversial field of imprinting. With imprinting and other so-called epigenetic modifications, genes can be reversibly altered within the lifetime of an individual, and new characteristics reflecting their own experience are passed on to their progeny.
An important part of any sequencing effort is the software used to assemble it from multiple, overlapping smaller pieces, and also later, to analyze it. Oxford has partnered with a company called Accelrys to develop a comprehensive analysis package known as Pipeline Pilot. It will tap into the huge set of bioinformatics algorithms already existing in the public domain. As those who have experience with DNA analysis from the smaller data sets available from companies like 23andMe know, there is no shortage of other powerful analysis programs out there.
Competition in these new markets is cutthroat. Life Technologies Corporation also intends to capture its share with its own sequencing device, and others are on the horizon. For the most part, these companies have completely ignored the DIY market, focusing instead on medical and research areas, and linking to the larger community only peripherally through genetic consulting companies, like Personalis. To some extent, new areas are already here; they are just waiting for a better sequencing technology. DNA-based computers and micromachines, at least as currently imagined, depend heavily on the ability to rapidly sequence long strings in parallel. Having devices like the MinION, and even smaller devices in the future — like those in the recently proposed Brain Activity Map project — will begin to open these new fields up to anyone so inclined.
Now read: The quest for the $1,000 genome