May 18, 2010 | Challenges and opportunities abound for Life Technologies, the new entity formed from the mega-merger of Invitrogen and Applied Biosystems. After temporarily ceding its monopoly of the sequencing market, Life is developing new technologies and strategies to win the near-term sequencing market and build an informatics and regulatory ecosystem for genomic medicine. Here, Life Technologies president Mark Stevenson tells Kevin Davies about Life’s plans for the present and future of genome medicine, and why he chose not to get his own genome sequenced.
Bio•IT World: Tell us about Life Technologies’ upcoming “third generation” single-molecule sequencing technology, based on quantum dots?
Mark Stevenson: What is Gen-3? There’s been a lot of discussion about single-molecule and real-time. [PacBio CEO Hugh Martin] says it’s Gen-2 plus… That’s not how we see it. We see the two techniques as complementary. High throughput is with the Gen-2 [SOLiD] or short-reads. We see Gen-3 has having capabilities of long runs and simple sample preps, but it has disadvantages, certainly today, of accuracy and throughput. So the two together might work, but at the moment, they haven’t supplanted one another.
Gen-4 is what we tend to describe Ion Torrent or IBM or things we’re working on and not ready to talk about today—non biological systems rather than using the polymerase. What if you go to an engineering based solution? That’s how we think internally. But just as Gen-2 hasn’t supplanted Gen-1 [Sanger sequencing], where we still do $600 million plus, we don’t think Gen-3 totally supplants Gen-2. That’s why we’re working on all three generations.
Is your strategy to maximize the performance of SOLiD and Gen-3?
Exactly, we see the two being complementary. Where you want high-throughput and counting applications, we’re the best in second-generation, to get to the $3000 genome by the end of the year and to go further than that in terms of accuracy. Gen-3 we see for applications that require long reads or very fast turnaround times. As we’re doing bases per second, we see that will get to a point where you’re in a hospital clinic, where turnaround time makes a difference.
How will your single-molecule Gen-3 technology compare to PacBio?
The system should be simpler, because we’re not forced to go to the wells and use nano-structures. So that should make it simpler for us to develop the instrument. The chemistry is portable, so we can tune it for extra accuracy and read length. That feature, we think, will be differentiating when scientists think about single-molecule sequencing and some of the parameters they want to choose in their instrument.
What about Gen-4 technology?
We see the application to continue to bring down the cost. Part of our job is to continue to scan and survey the technologies out there—we have feasibility programs looking at those technologies. We think in time there’s a way to go in solving those problems. We have the largest sequencing business in the world, so we have to continue to look at those technologies.
Do you see SOLiD making inroads and closing the gap on Illumina?
I think the fact we came in later to the market meant that for those with an install base, there needs to be a reason to switch. In a case like [the Ignite Institute], they could compare both platforms without an install base, they could leapfrog some others. The advantage we see is first accuracy. We took the advantage we have in color space, and are further developing that advantage. The informatics to deal with color space is harder. So we’ve developed a way to read out color space in base space, without reducing accuracy. That translates in applications like cancer, where accuracy matters. Typically with Illumina, we’re seeing error rates of about 1% [99% accuracy]. We’re getting 99.94% accuracy. That makes a big difference.
During this year, we’re increasing [2-base encoding] to 5-base encoding. That advantage in accuracy will go even further in terms of base calling. By the end of the year, we’ll be up to 99.99% accuracy. Ultimately, we think the equation is accuracy times throughput—we’ll be at 300 Gb—divided by cost. We’re at $6000 [per human genome], and ultimately we’ll be at $3000. We believe ultimately that’s what people want—that’s the quality genome, that’s what we see is our advantage. We’ve taken a share of the market, but we think the market is still getting going.
One criticism against SOLiD is the DNA sample prep requirements. Have you addressed that?
We’re aware of that limitation in the early versions. We just launched the EZ Bead system to automate the upfront sample prep. It reduces hands-on time by 90% and some of the variation in the sample prep.
Apart from sequencing, what technology areas are you most excited about?
We just introduced a flow cytometer, which is also a very disruptive technology, using acoustic focusing to separate cells based on size using sound waves. It’s very innovative and disruptive. One of the things Life Technologies wants to do is lead with the science. We’re all about transformative science, so we do want to enter transformative areas…
The explosion in informatics and genomic information is fascinating. Where it links back to bioinformatics is in two areas. One is this whole sequence-to-function—ultimately you do need to get at the functional side. That means internally either validating using real-time PCR or protein-based techniques. There’s a lot of work in our cell systems division, where what we’re doing is to link the genomic information together on our web site, so you can conveniently order, having identified the gene, the RNA to knock it out, the protein or Taqman assay to address it, or a cell-based assay to go from sequence to function.
Second, how do you ultimately present this information to a physician making a decision? That’s where we want to go next. In the next couple of years, we will make a $100-million investment in IT: how can you turn this vast amount of data into knowledge? We look at our road map, the technology is getting to the $1000 genome, but how are you going to call it, to present it to a physician? We’ve started various partnerships with various IT companies, but where do we go next? Life Technologies has the scale to think about this and execute.
It’s more about genome visualization and informatics. One of the models we’ve looked at is, if you look back into the past like radiology, when a physician was capable of looking at a scan and saying this is what’s happening with this patient, over time that evolved into CAT, PET, imaging, and a more digital age… What’s the equivalent in genomics? It’s the sequencer that is now spinning out information. What does that look like when you want to read information? Also, who is going to read that information? Two parts to it, that part is more around a genomic doctor, equivalent to the radiologist. We’ve also set up $5 million to fund the creation of a sub-specialty of genomic medicine doctors to read this information.
You’ve announced a partnership with the Ignite Institute, which is acquiring 100 SOLiD instruments. What prompted that partnership?
We hope to get methods and protocols on how you do large-scale sequencing on center patients on diseases such as autism, and use that information to develop associated protocols associated. Second, we really want to understand the regulatory framework. So the fact [Dietrich Stephan] located it in the DC area allows him to engage with the regulators as a proof of concept. And we’d like to see how that’s going to come forward in the health care system.
Life Tech CEO Greg Lucier reportedly had his genome sequenced. Have you opted to have yours done as well?
No. My main reason was, what extra information do I get out at this point about the genome, compared to spending that money and sequencing a group of 14 breast cancer patients? So it was a personal reason: how do we spend that money and for what reason?
Last year, we paid for all the senior [Life Technologies] leadership to have a Navigenics scan. Interestingly, 50% didn’t get it done. Some of that was apathy, some was concerns on privacy, because it was the company. Some didn’t see any value to it. I didn’t find anything ‘useful.’ There was no elevated risk... For someone like Jim Lupski, who had a genetic disease, it’s worth sequencing to find the rare variants. But I didn’t find there was a reason to do it.
Greg had his genome done, and Kevin [McKernan] compared him to Craig Venter! So you discover the SNPs, but what do the other SNPs mean? You need a physician, like Daniel von Hoff, to advise on why you’re doing sequencing. That, for me, is the missing why. •
This article also appeared in the May-June 2010 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.