Reading the sequence of DNA in a human genome cost about $1 million in 2007. Now it costs $10,000 to $20,000. The ability to cheaply sequence large volumes of DNA has already led to the discovery of new disease genes and improved our understanding of evolution and human history. But Jonathan Rothberg, founder of Ion Torrent, says that DNA sequencing is on the brink of another transformation—and this time its impact will be much more personal.
Rothberg likens the state of genomic technology to that of computing before PCs. And in much the way that the microchip made personal computers and smart phones possible, Rothberg predicts, Ion Torrent’s small, cheap machine will allow sequencing to seep into medicine, agriculture, and energy.
Ion Torrent’s tabletop machine costs $49,000, about a 10th as much as other “next-generation” sequencers. It uses a disposable $250 chip that is fabricated in microprocessor foundries. Industry leader Life Technologies acquired Ion Torrent for $375 million in 2010.
Rothberg recently told TR’s biomedicine editor, Emily Singer, why he has such high expectations for his product.
TR: How will this technology transform medicine?
Rothberg: This is really about creating a personalized medicine. Our sequencers are for research use only now, but eventually we see them being placed in all labs and in any clinical setting. That means doctors will be able to use genetic information to make decisions, such as selecting medicines.
Since it is a 10th the cost of other technologies and 10 times as fast, it will be an ideal fit for medical settings, such as clinical genetics labs and pathology labs. And it works in a time frame of hours, which matches the decision-making time frame that physicians often have.
Where do you think the technology will be adopted first?
We are already seeing it put into use for cancer and infectious disease. Researchers at Massachusetts General Hospital are setting up a system to look at 200 hot spots [regions of DNA that have been linked to cancer] in tumor samples, which could be used to determine potential treatments and to predict outcomes.
People have been talking about incorporating genetic testing into medicine for years, ever since the Human Genome Project. What’s different now?
This is absolutely the turning point. Patients will be just as likely to have their genomes sequenced as they will be to get MRIs or CT scans. And sequencing doesn’t just tell you where you are; it can tell you the future. You can see whether you are predisposed to disease.
How does it work?
Our chip is laid out like an imaging chip in a digital camera, with millions of sensors that directly detect changes in chemical signals. Over each sensor, we fabricate a little well. Each one of those wells is an independent sequencing machine. One piece of the DNA of interest goes into each one of those wells. Every time a base [or DNA letter] is incorporated into the growing strand of DNA, it releases a hydrogen atom, which the sensors detect. It’s kind of like the world’s smallest pH meter.
Other machines on the market can analyze more DNA than yours. Isn’t that a problem for Ion Torrent?
When you look at our personal genome machine, you are looking at the equivalent of the first video game. Think about what Pong looked like, and then think about what Xbox looks like. In the next few years, you are going to see this transformation [in sequencing].
Right now, we can fit 40 machines [meaning the individual sensors that detect DNA letters] on a human hair. If we make chips in a new foundry, rather than the foundry we are using now, we can fit 4,000 machines on a hair. And we can go even denser. Memory chips have billions of cells, and there is nothing stopping us from doing the same. We plan to make sensors that can directly detect billions of simultaneous sequencing reactions.
You haven’t yet sequenced a complete human genome—does that matter?
We found out that 80 percent of labs today want to sequence sets of genes, from one to 500 genes—not exomes [the portion of the genome that corresponds to genes] or genomes. The first chip is designed for that market. The diagnostic samples are huge opportunities. [Clinical labs] do 25,000 to 100,000 samples in which they want to sequence 10 to 500 genes. For HIV, you sequence [particular] viral genes a million times to see if the virus has grown resistant to certain drugs.
You have founded a number of sequencing and genomics companies. What was your guiding principle for Ion Torrent?
We took the same approach as the computer guys. Rather than a large, expensive instrument that is hard to build, ship, and set up we aimed [for something more like] a personal computer that anyone could set up and use.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.