The Gadget that Makes DNA Sequencing Child’s Play

The MinION cracks open biotech to the masses the way the PC democratized computing. What will we do with this newfound power?

The MinION (Courtesy of Oxford Nanopore)

I t is a Tuesday afternoon and Poppy, a 12-year-old girl in New York City, stands in front of her class and explains to her peers how the code of life can be read by passing a DNA strand through something called a nanopore. As part of PlayDNA, a program I co-founded, the students have been pickling cucumbers for the past week. They’ve measured the pH of the liquid in the pickle jars and saw from the increasing cloudiness that the number of the bacterial cells was doubling. And unlike generations of science classes before them, they’ve taken samples from the jars to identify the bacterial species by their DNA.

It’s now time to reveal the invisible life in their pickle jars. The students gather around the table and, together with their teacher, put a real bacterial DNA sample in a tiny DNA sequencer, which simply plugs into a computer’s USB port. Minutes later the first DNA reads appear in real time on their screen.

This is possible in a middle school because of the miniature DNA sequencer, called the MinION, made by Oxford Nanopore Technologies. I’ve been using this device for nearly two years at the New York Genome Center, where I research how to use it for re-identification of DNA samples. My adviser, Yaniv Erlich, and I were the first to implement it into a Columbia University classroom, and now it’s part of our PlayDNA program in local schools. I’m convinced that it represents a milestone in technology. Portable DNA sequencing empowers anyone, not just scientists, to see life at a higher resolution than the fanciest camera can provide — and even after a creature is gone. We can broaden our vision to see all species, not just the ones that are visible to the naked eye.

The MinION costs $1,000 and is the size of a candy bar. It connects to a laptop computer’s USB port. To have it read a DNA sample, you use a micropipette to drop a “DNA library” (more on that in a minute) through a millimeter-sized opening on the MinION. Inside the device are nanopores, cones just over a billionth of a meter wide, placed in a membrane. A steady ion current flows through these nanopores. Since each nucleotide (A, T, C or G) has a unique molecular makeup, each one is shaped a little differently. The unique shape passing through the pore interrupts the ion current in a specific way. Just as we can infer a shape by analyzing its shadow on a wall, we can infer a nucleotide’s identity from the disturbances it causes to the ion current. This is how the device converts bases to bits that stream into a computer.

An illustration of how DNA and a current flow through a nanopore. (Courtesy of Oxford Nanopore)

We are not yet able to directly micropipette pickle juice into the MinION. Some advanced steps are required to prepare the DNA library that is sequenced. First you have to crack open the cells in the pickle juice and purify their DNA. Cells are all different — you might recall from biology class that plant cell walls look unlike bacterial cell walls, which are unlike the membranes of mammalian cells — and each cell type requires its own method. Then, the purified DNA needs to be prepared in such way that the MinION can actually read it. These steps to create the DNA library require machines that are not yet user-friendly for a non-specialist, including a micro-centrifuge and thermo cycler (at Democratizing DNA Fingerprinting you can see me performing this library prep and DNA sequencing on a rooftop in New York City). But in the future, these steps will also be done in a single, portable miniature device.

This will open up the field. People will be able to use the MinION in their kitchens to verify the contents of their ready-made lasagna (does it really contain beef or is that horsemeat?) or use it for surveillance of pathogens and allergens. Oxford Nanopore is even planning to go one step further with the SmidgION: a DNA sequencer you can plug into your phone.

But we’re still only beginning to see what people will do with this technology. Scientists have taken advantage of the MinION’s portability to monitor biodiversity in remote areas such as Antartica’s McMurdo Dry Valleys. NASA is using the device to monitor astronauts’ health status in space and could eventually use it to visualize extraterrestrial life. Authorities in Kenya might soon check instantly whether meat comes from illegal poaching.

In our lab at New York Genome Center we developed a method to use the MinION at crime scenes. We figured that a portable sequencer, which can deliver results in minutes, could give investigators a head start on identifying victims or suspects. Traditional forensic methods can take days, sometimes weeks. That’s because someone has to transport the samples from crime scenes to well-equipped laboratories, where the evidence sits in a queue before being run though expensive machines.

Nanopore sequencing sensors are an addition to the genomics field and are unlikely to replace the more traditional sequencing platforms, like those produced by the market leader, Illumina. Those DNA sequencing platforms are extremely accurate, making them indispensable for reading an entire genome (a couple of times), which is what is needed to, say, determine which genetic variations in people lead to diseases.

That kind of work isn’t currently the strength of the MinION. It has an error rate of roughly 5 percent, which means that there is one reading error every 20 nucleotides. That’s high considering that the difference between two individuals is 0.1 percent (one variation every 1,000 nucleotides). But the readout from the MinION is still good enough to feed into the algorithm that we developed for crime-scene analysis. This algorithm computes the probability that hair or some other material found at a crime scene matches an individual in a special police database.

To understand why this works even with the high error rate, imagine that I give you the name “Voldamord” and ask you to tell me what book I’m referring to. You might recognize it’s a Harry Potter book because you have a database in your head that has been formed through reading, even though there are typos in the word I am giving you. You don’t need to re-read the whole 300-page book or get “Voldemort” presented exactly right. Genomics works along the same principle. Once you have a useful database, you need only some informative DNA fragments to identify which bacterial species are present in the pickle samples or sometimes even which person the DNA came from.


Now that the era of ubiquitous DNA sequencing is getting closer, we need to improve genetic literacy. How do we handle this genomic “big data”? To address such questions, Yaniv Erlich and I started a class called Ubiquitous Genomics in the Columbia University computer science department in 2015. We taught students about this cutting-edge technology and got them to experience the potential. The students sequenced DNA with their own hands, and were encouraged to develop computational methods to analyze their data. The success of this effort in “integrative learning” encouraged us to think we could do something similar to engage schoolchildren in genomics and data analysis. We founded PlayDNA with that aim.

A closeup of the micropipette used with MinION. (Courtesy of Oxford Nanopore)

The day before the start of the first PlayDNA pilot class, I set apart a couple of ingredients from my lunch that would later end up in a mystery DNA sample the students had to identify. PlayDNA provides the infrastructure for classrooms to not have to worry about extracting DNA and preparing the DNA libraries, so students can start sequencing DNA right away and interpreting their data. Twenty 12-year old students, who got only a couple of hours of micropipette training, were sequencing DNA not two hours after arriving in the classroom. The real-time conversion of biological information into big data enlivens the subject; the students were eager to know which species could be spotted in the DNA readouts they were seeing. Their assignment for the following week was to analyze the data and identify the ingredients and their ratios of my lunch. Sure enough, the following week one group asked: “Sophie, did you eat a tomato salad and some sheep meat for lunch?”

Is the technology ready for your kitchen counter? I would hold off on making space for a while. It still takes some know-how to handle the steps prior to sequencing, like breaking the cells open and purifying the DNA. Oxford Nanopore is working on ways to automate these steps as well, however. Eventually, I can foresee a family where the kids are using a SmidgION to play a new version of Pokemon Go in the park with real species, while mom asks dad: “Darling, did you set the table and did you sequence the lasagna?”

Sophie Zaaijer is a postdoctoral fellow at the New York Genome Center and CEO of PlayDNA, which is developing genomic-data classes for middle schools, high schools, and university education.