Name: Dietrich Stephan
Title: Founder, president and CEO, Ignite Institute for Individualized Health
Experience and Education:
Co-founder and chief science officer, Navigenics, 2006-2008
Deputy director for discovery research and founding chairman, department of neurogenomics, Translational Genomics Research Institute, 2003-2008
Senior staff fellow, National Human Genome Research Institute, 1997-1999
Postdoctoral fellow, NHGRI, 1996-1997
PhD in human molecular genetics, University of Pittsburgh, 1996
BS in biology/biochemistry, Carnegie Mellon Institute, 1991
Two weeks ago, Life Technologies said that 100 of its new SOLiD 4 sequencers will go to the Ignite Institute for Individualized Health, a new research institute in Northern Virginia, which will thus join the ranks of other large US genome centers (see In Sequence 2/2/2010).
Earlier this week, In Sequence spoke with Dietrich Stephan, Ignite's president and CEO, about how he wants to apply this sequencing power, which will be installed over the course of the year.
Stephan, who spent several years at the Translational Genomics Research Institute studying genetic links for a variety of diseases before co-founding personal genomics firm Navigenics in 2006, founded the institute last year. Below is an edited transcript of the conversation.
Can you briefly provide some background about the Ignite Institute and its mission?
It's a new non-profit research institute focused on personalized medicine. We hope to put in place technology that allows us to sub-classify common diseases at the molecular level, and use that as the starting put for new exposure studies and new drug development studies.
[ pagebreak ]
Why is it located in Northern Virginia?
We wanted to put it into the national capital region because we needed to have access to policy makers and regulatory bodies, so we could get these new, complex signatures and drugs approved. We basically wanted it to be near Washington, DC.
What is the current status of the institute? How many scientists have you recruited, and where will it be housed?
We are just starting operations. Right now, we are in the process of recruiting scientists to come here. The institute will be housed at the Center for Innovative Technology near Dulles Airport. That's temporary, and then we will be moving into a permanent facility in about a year, and we have not identified that yet.
How much funding have you raised so far, and from whom? How is the institute going to be funding its operations?
We have raised about $250 million in funding to start the institute, and that's about halfway to our five-year goal. It's a large public-private partnership involving the state of Virginia, Fairfax County, the local healthcare system — Inova Health System — and then we have some university partners, as well, who together make that up.
[Going forward, the institute will be funded] through traditional academic grants, contracts, donations, and intellectual property revenue.
You just announced that you will install 100 SOLiD 4 instruments this year. How did the deal with Life Technologies come about? Did you consider other platforms as well?
We did. But we really like the [SOLiD] technology, and we really liked Life Technologies as a company, meaning that the corporate mindset matched our institutional mindset very nicely. They really wanted to make a difference in healthcare and would be a great partner for us in the long term.
What does the partnership with Life Tech entail, apart from the company providing you with instruments?
We really hope to, together, be able to develop new ways to use the genome sequence to develop new diagnostics. How do you move from a machine in a clinical laboratory to something that you can do for a patient that can improve their health? It's that information space that we are going to be working on together. They want to really understand, like we do, how the information will be used in the future.
[ pagebreak ]
Did you consider outsourcing sequencing to a service, like Complete Genomics, and why did you decide against that?
We have really wanted to be able to control the process and control the queue. It's just a matter of being able to do the projects we want to do in the timeframes we want to do them. We did not want to be reliant on a third party.
Apart from the SOLiDs, are you looking into some of the newer, long-read technologies as well?
I think we are just going to stick with what we have for now and try and get up and running with these machines.
What does it take to set up a genome center on that scale from scratch?
I think we are going to learn that. I have certainly never done it before. We will phase in the machines over a year, and learn from our colleagues in the scientific community [about], for example, the information technology side and the data storage and all of these issues. We view this as an asset for the scientific community.
In what sense is it going to be an asset for the scientific community?
Really, being an open collaborative framework. We are going to try and work with everyone and anyone to leverage this asset. It's something we did before with my group at TGen: We had this microarray consortium where over a 10-year period we were able to allow 2,000 investigators from across the country to come in and access the infrastructure, generate data. We would work collaboratively to make sense of the data. So the notion was that they would not have to go off and build a big array center. We want to do the same thing here; we want to try and be highly collaborative, so that as a community, we can get some inroads into these diseases.
Will the sequencing center be CLIA-certified? Why would that be important for a research institute?
A portion of our facility will be CLIA-certified. We definitely are a research institute, but we are also translational in nature. So once we develop something, we want to be able to turn it on in a clinical setting, and the CLIA infrastructure will be important for doing that.
[ pagebreak ]
What kinds of studies are you planning to apply your sequencing capacity to?
On the research side, we want to go back and redo all of the whole-genome association studies that we have done over the last five years or so. These are studies my group did [at TGen] — we did about two dozen whole-genome association studies. We will go back and redo all of those samples using a full-genome sequencing strategy to hopefully extract the remaining heritability that we haven't found yet for these diseases.
On the clinical side, we would like to start with cancer and try and see if we can't apply some of these newer technologies at the point of diagnosis of the cancer, so that we can better guide physicians to which of the drugs they should use in the standards of care — so stratifying the standards of care — and then also try to develop experimental off-label strategies for patients once they progress.
Where are the patients going to come from? Are you collaborating with a cancer hospital for this?
Right now, we are thinking about [those] coming from our local Inova Health System. But we will work with a ton of different clinical partners.
Will the GWAS studies largely exhaust your sequencing capacity?
That will keep us certainly busy for a couple of years. Even with 100 machines, we will only be able to sequence 7,500 genomes a year. That's just a couple of these types of studies every year. I think as the technology continues to evolve, we will get higher and higher throughputs. But we really just wanted to get started on a couple of key diseases, including Alzheimer's, Parkinson's, autism, bipolar disease, and multiple sclerosis.
Is our knowledge of the genome today advanced enough to make sense of whole-genome data? A number of other groups have opted to start with exome sequencing.
We know from the whole-genome association days that there is important information that's outside of the exome. That's why we want to just go right to sequencing the full genome.
In terms of interpreting, I think it's just a data reduction and filtering process that happens. I don't think you need to, on the first day, look at every single one of the 6 billion letters, I think you can go through and say, 'What's different relative to the reference sequence?', filter that, and then, instead of 6 billion bases, we are down to 30 million, or whatever it is. And then, suddenly, you start to be able to look between cases and controls more effectively and just see what's different, and then look for clustering of variations in certain regions between the two cohorts.
That will give you very rapidly a high level of understanding of whether something weird is going on in one region, and once you have that, you go to the next level, which is, 'What are the functional consequences of the variations, do they change amino acids, are they in conserved regions?' I think you should be able to get a pretty good understanding if you have a bunch of variants, whether it's a loss of function, or a gain of function, or change of function-type of mechanism. If you have variants that are clearly deleterious that are clustering in a gene, then you know it looks like this gene's function is turned off, without going into the details of every single variant. Clearly, we haven't done this before. This is my thought process of how this will pan out.
[ pagebreak ]
Will the institute also have other research areas besides genomics — for example proteomics, cell biology, or systems biology?
I think we are going to put the technology in. We are going to try and lead further in the technology side to be able to study DNA, RNA, proteins, and other metabolites.
And then we are going to try and partner on the analysis side. We have already been working with Virginia Tech, and Skip Garner there [the executive director of the Virginia Bioinformatics Institute], who will help us analyze this, but also the Sage Bionetworks group. I really see us making our data publicly available and then pushing it to groups like Sage where Eric Schadt and Stephen Friend can help understand what it means.
Are you planning to partner with companies for commercializing research results?
That's the goal, to do co-development with companies like pharmaceutical companies, or license our IP out, or start a new little company.
Are you still involved with Navigenics?
I am. I sit on the board. It remains a really great company — we have a new CEO, Vance Vanier, who is spectacular. The business model of Navigenics is really to take the public domain information and synthesize it and then apply it to a person's genome in the form of rules. In that respect, as we put data from our research institute into the public domain, it will get sucked into Navigenics just like all of the other information.
What are going to be the first areas of medicine where individuals will benefit directly from genomic studies you will be conducting at your institute — which is called, after all, the "Ignite Institute for Individualized Health"?
I think it's going to be cancer in the clinical research setting, and probably Alzheimer's disease after that.
Why Alzheimer's?
The genetics are worked out pretty nicely for late-onset Alzheimer's disease now, and we are starting to get to the point where there are some promising subtly effective primary prevention strategies coming online, like mid-life cholesterol levels and things like that, so that we can start to think about identifying genetically [at risk] individuals and giving them something that might be effective as a long-term prevention strategy.