On the application spectrum of the PacBio RS system: Targeted resequencing in humans and in diseases such as cancer is important. An extremely important area for our customers is studying transcript structure. In the near future, we’ll be looking at chemical modifications of DNA including the very broadly studied methylcytosine, the less well understood hydroxymethylcytosine, and also the broad range of other modifications that are present. We’ll also be looking at hybrid assembly, finishing of genomes through the combination of data from the long-read third-generation PacBio system and the shorter-read but higher-throughput second-generation systems.
Another important area is infectious disease. We’re getting extremely good traction with the Centers for Disease Control and the FDA, especially following the proof of concept we published in the New England Journal of Medicine regarding the cholera outbreak in Haiti. Our system, because of the fast time to result and the granularity with which we provide sequence, is ideally suited toward monitoring, diagnosis, and a number of other applications in the infectious disease area. The long read length is an extra bonus that allows us to see certain types of modifications in highly divergent populations of viruses and bacteria that some other platforms can’t. So those are the four main areas of interest right now. As the raw read accuracy increases, which it is doing at a fairly continual pace, that may expand to include de novo sequencing using the PacBio system only. And of course as we increase the throughput, the applications in those four areas become more economical as we move forward.
On the absence of target amplification when using the PacBio RS system: Our perception is that the primary differentiators right now are the read length, the time-to-result, and the granularity that are driving the application. The lack of target amplification is an important feature, though. For some applications it’s enabling. For example, take detection of chemical modification. We recently presented results on the detection of hydroxymethyl cytosine, for which target amplification wipes the signal out. In these contexts you can’t really say it’s an attractive feature; it’s an absolute prerequisite for many of the kinds of things that people want to do. For the other applications, it’s certainly attractive because it reduces the complexity of the upfront sample preparation. It reduces the time; it reduces costs; and those are all attractive features. We certainly do envision that at maturity this technology will be able to produce results from unamplified DNA in a very short period of time.
There are, however, important applications that do involve target amplification. For example, at the AGBT conference, Gen-Probe presented data from our system showing the ability to detect novel variants of the HCV virus. And there they used Gen-Probe’s proprietary and rapid transcription-mediated amplification.
On the future of nanopore sequencing and the possibility of it becoming a commercial reality/competition: I think that the experience PacBio had is a good gauge for understanding how long it takes to get a completely new technology into the marketplace. Ours was not actually that different from the Illumina or 454 Technologies in terms of the time frames from the first sequence of a short oligonucleotide to when an instrument can be shipped, and so forth. And it’s always measured in years. That means that when we see a nanopore technology that produces that first sequence of DNA, we can expect it will be at least several years before that technology is ready for the commercial marketplace.
Of course PacBio is watching new technologies very carefully, and at some point the zero-mode waveguides may give way, but our commitment to enabling our customers through longer reads and cheaper runs will not give way. We intend to remain in the forefront of sequencing through the various generations of technology that will come and go.
I think nanopore as it is conceived by many people with a direct readout may never happen. But I think it’s reasonable to think that there will be techniques other than phospho-linked fluorescence-based nucleotides with DNA polymerases. You’ll notice that what are turning out to be the most feasible techniques using a nanopore to sequence are drifting ever closer to the PacBio SMRT sequencing concept where you have a polymerase molecule and you’re using the nanopore to eavesdrop on the activity of that polymerase molecule. So I think that beyond any doubt generations of technology come and go, and while it’s hard to imagine exactly what those technologies are going to look like I think we can be certain that they will come.