Andreas Sundquist is the CEO of DNAnexus, a cloud-based DNA sequence storage and analysis company. As the sequenced human genome becomes commonplace, the amounts of data involved grow to staggering proportions. DNAnexus wants to be the repository for your genome (and it’s a near-certainty that yours will be sequenced in the next decade), in the brave new world of personalized medicine. Having garnered a $15 million in funding last fall in a round led by Google Ventures, DNAnexus plans to launch a revamped platform–one focused on collaboration among researchers–later this year.
FAST COMPANY: What is DNAnexus?
ANDREAS SUNDQUIST: DNAnexus is a DNA data management and analysis platform. We believe it will power the genomics revolution. The goal is to unlock the potential for DNA-based medicine with a really collaborative and scalable data platform.
How is it different from a company more readers may be familiar with, 23andMe?
23andMe’s approach is to enable consumers to look at genomics. We look at it more from a medical and biotech angle. Ultimately we see genomics being used in the clinic, with diagnostic labs using DNA sequencing to help you understand your health, whether a certain drug works for you, and whether you’re at risk for a certain disease.
Are there applications of DNA sequencing beyond medicine?
There are a lot of applications, including agricultural tech and bioenergy. Bioenergy is the application of biology to improving energy production. For example, there’s a lot of excitement around algae that harvest sunlight to produce sugars that can be turned into energy. Genomics could be used to improve the ability of those organisms to produce energy more efficiently.
Your own background is in computer science. How’d you get into DNA?
I saw there was a really dramatic revolution of DNA sequencing technology. Ten years ago it cost a billion dollars to sequence one human genome; now it costs a few thousand dollars, and in the next five to 10 years, it’ll be cheaper than a lab test. That means we’re going from a technology that nobody had access to to a technology that billions of people are gonna be using. Here’s where it becomes interesting for us as a company. One human genome is almost a terabyte of data. Now multiple that by a billion genomes, and it’s just a tremendous data problem.
What, on a kind of CSI level of scientific sophistication, is it that makes computer science useful here?
Take someone with cancer for example. If we get DNA of your normal cells, and then get DNA of your cancer cells, and then you line them up against each other to compare the differences, you can see that a mutation is in a particular gene, and then you could tell the cancer patient, “Take this drug instead of that one, because we know it targets that gene.” But there are 6 billion letters of DNA in your cell, and 6 billion in the cancer cell. Using human labor, that’s impossible; you need a computer to churn through all that data.
You’re based in Mountain View, like Google, which has given you backing through Google Ventures. 23andMe is also based out there, and has close ties with Google (its CEO is married to Sergey Brin). What is it about Mountain View that makes it a genomics mecca? Are there DNA cocktail parties?
I don’t know. Maybe there’s something in the water here. But Silicon Valley and Mountain View are kind of a mecca for all of tech. Our relationship to 23andMe is cordial. We know them, but we don’t work directly with them right now, and we plan to be in very different but complementary spaces. I don’t know very many cocktail parties that are specifically for DNA people, but that’s a great idea. Maybe we’ll start those up.
You have a number of competitors–Illumina is one. How do you differentiate yourselves?
There are companies that overlap with us. The competitive space out there divides roughly into two groups: one is the traditional biotech or bioinformatics companies, but those tend to be missing people with a really strong sense of computer science and software scalability. The other group are the big data companies out there; we’re talking about petabytes or exobytes of data, and it’s conceivable that some of those companies will find the sequencing or genomics space attractive. But what we uniquely bring to the space is a deep understanding of genomics, how the whole ecosystem ties together, and how to build tools that a clinician really wants to use. We have a unique combination of both the computer science and the genomics understanding.
What is your business model?
Pay-as-you-go. Our pricing is proportional to the volume of data you use. One price point, as an example: we charge $10 per gigabase for an academic. A gigabase is a billion letters of DNA, and included in that is all the storage, tools, and visualization, bundled into one package.
You talked about how far DNA sequencing has come in the last 10 years. Where will it be in the next 10 years?
It’s going to be cheap as a routine lab test–a few hundred dollars. In the next ten years, you’ll see every single person in the developed world having their genomes as part of their record. As soon as you’re born, you’ll have your DNA sequenced. That’s a world of change. If you want to run a diagnostic test in the future, you can really just run software on data that’s already there. The ability to personalize treatment is something that’s right around the corner, and I think DNA sequencing is one of the keys to enabling that.
This interview has been condensed and edited.