Interview: Jonathan Rothberg on speed-reading your genome

In an interview he gave to the journal Nature last year, Jonathan Rothberg, the CEO of biotech company Ion Torrent, cited Steve Jobs as his biggest influence. While that's probably true of many tech entrepreneurs, Jobs recent death, from cancer, is bound to have affected Rothberg more than most. That's because the Connecticut-based engineer and serial entrepreneur invites comparison to the former Apple CEO in a way that few people do. After all, just as Jobs revolutionized personal computing, Rothberg is doing the same for biology and medicine. At the heart of both revolutions is the humble silicon chip.

About 10 years ago, Rothberg pioneered a faster and cheaper method for reading genomes called next-generation sequencing, which is currently the gold standard in research labs around the world. Now, he has launched a desktop gene machine that may finally usher in the long-awaited personal genomics revolution by dramatically cutting the cost of decoding an individual’s DNA sequence and fingering their genetic weaknesses. This, in turn, creates the possibility that we'll soon be able to diagnose and treat a host of diseases on an individualized basis -- chief among them, cancer. Unlike its predecessor's, Rothberg's new invention -- the Ion Personal Genome Machine (PGM) -- reads DNA using semiconductor technology, making it cheaper, faster and more scalable than any other.

Due to the sequencing power of both generations of his machines, Rothberg has laid claim to a lot of firsts: he led the effort to sequence the first individual genome (James Watson’s of the double helix), initiated the first large-scale sequencing effort of ancient DNA -- the Neanderthal Genome Project, and helped crack the mystery behind the massive disappearances of the honey bee, as well as a deadly E. coli outbreak in Germany.

We sat down with Rothberg at PopTech 2011 to discuss how making DNA sequencing more accessible stands to transform medicine.

PopTech: You've sequenced the genome of James Watson, one of the co-discoverers of the structure of DNA. Have you sequenced your own?
Jonathan Rothberg: I get this question from my wife because I recently sequenced Gordon Moore's genome, who's the founder of Intel. And, as you mentioned, I’ve sequenced Jim Watson's genome. She asked me, “Why do you always sequence 80-year-old Caucasian men? They’re healthy.” I sequence them precisely for that reason. Because except for educating people about why it’s important to sequence for medicine, for discovery, for making drugs, for diagnostics, for understanding the progression of disease, for finding a cure for breast cancer, I think genetic materials is private and that you should have a reason to sequence it. You should be sequencing because you are trying to understand disease, you should be sequencing because you are trying to make a diagnostic, you should be sequencing because you are making a drug. So, no, I haven’t.

What's your vision for the future of genome sequencing and personal genomics? Some scientists have suggested that every baby should have its heel pricked and its genome sequenced at birth.
My vision is that sequencing will develop in an analogous way and be equal or greater in importance than imaging has been to medicine, just as how part of medical practice we have X-rays, MRIs and CAT scans.

I do, though, have a vision that starts with the heel prick, where, in a newborn unit of a hospital, every child has his or her sequence done. And I think there will be a time when that will make sense -- when the economics makes sense and when we have data that correlates sequence with disease, sequence with things we can take action on. Then it will make sense to sequence the whole genome.

Over the next five to 10 years, we will have to be sequencing -- and working to make sense of the sequence -- so that a decade from now, when that heel is pricked, we'll be able to do something. For example, does that child need a different diet, or should that child stay out of the sun? So right now, I think sequencing is best done as it's needed -- so you have a person with cancer or a newborn who is sick and you use that sequence to inform medical decisions. As we have more medical information along with the sequence, I think it will become a more general tool. And the nice thing is that by that time, it will be cheap enough that it can be universal.

So do you see a personal genome machine in every doctor’s office?
I do see it there in a decade -- and maybe I'm biased because I'm married to a physician -- I do see it in the context of normal medical care. On the back side, I have a cousin that got a gene sequence back for Huntington’s chorea and she killed herself. So I know it has to be done in the context of medical practice, just as it would be when you give someone an AIDS diagnosis. When I have an MRI, or an MRI of my child, it's not emailed to me. I sit down with a radiologist who knows the family history and knows the reasons why the MRI was taken and puts everything in context. And that's the way genetic sequence has to be used. Whether it’s a single gene sequence, whether it’s 300 cancer genes or a whole genome, it has to be done in the context of medical practice, with expert counselors, support and guidance.

We’ve got a lot of work to do to catch up with the technological advances…
The good news is that while I’ve been focused on technologies, there are people who are focused on making sure we do this right: that [genetic information] is not used to not give people insurance, that it’s not used to not hire somebody. So, as I go to work to make sequencing cheaper, there are other people working to make sequencing part of medical care, and to take the necessary precautions, put in place the privacies and establish the guidelines that must go hand in hand. 

Do you know much it costs to sequence a genome today?
I know precisely where we’re at. If you do genomes now, they are on the order of $10,000 to $20,000. By this time next year, we’ll be doing them internally -- meaning within the companies that are developing the technologies -- for less than $1,000. The price is dropping faster than Moore's Law because we are building on 40 years of technology and catching up quickly. So by 2013, it will be less than $1,000.

One thousand dollars has been a magic number in the developed world because in the United States we spend $6,000 per person per year on health care. So you can image that if someone is going into a newborn intensive care unit, a $1,000 expense looks pretty attractive because it might save $50,000 in healthcare costs in a week's time.

And it will keep on getting cheaper. I can predict that within a decade, it will be as cheap as that heel prick. A heel prick and all of the related biochemistry costs about $10.

Can you tell me about your new technology -- the Ion PGM sequencer that you used to sequence Gordon Moore's genome? How is it an advance on previous technologies?
I'll use an analogy. Next-gen sequencing was all about sequencing fast, in a massively parallel way, in the miniature. But it used light. If that generation of sequencers, that next-gen, was equivalent to the large computer centers that allowed universities to compute, then Ion Torrent is analogous to a personal computer.

We stepped back from the problem, started from scratch and said, what we want to do is make a chip that sees chemistry, completely analogous to a chip in your cell phone that allows you to take a picture. That kind of chip takes photons and turns them into electrons and turns into a computer image. We now have a chip -- again, analogous to the chip in your cell phone -- but instead of seeing photons, it sees chemistry. We literally just put the DNA on the chip and it reads it off. So it’s the first time that you can sequence without light, which makes it much more efficient. And it’s the first time you can sequence by taking advantage of all the technology that gave us modern, inexpensive electronics. So now we can truly democratize sequencing.

When we introduced our machine, it was 10 times cheaper and, you can say because of that, it’s the best-selling machine. But that made a difference. This summer there was an outbreak of E. coli in Germany and people were dying. Because our machine was inexpensive, labs in Germany and in China had it. Because our machine was about 100-times faster than the generation before it, within two hours, people were able to decode that bacteria and put the sequence on the Internet. That was really a game-changer. Scientists around the world looked at it and deciphered it and were able to say what this new bug was. And they were actually able to make a quick diagnostic kit that only cost one dollar to be able to check food sources. Now, if it ever comes up again, we can fingerprint it.

Photo: Kris Krug for PopTech

Rate this post:

  • Meh.
  • Love it!
  • Community Rating: 8
Click and drag above to vote.