A Hammer Thrown at Big Brother?

[NextGenSeq]

Oxford Nanopore Technologies, the company that pioneered a new DNA analysis technique called nanopore sequencing, has remained secretive about the microscopic channel at the heart of its products. But on the heels of a patent infringement lawsuit from sequencing behemoth Illumina, Inc., the company has revealed what will drive its newest devices—a bacterial pore that seems to circumvent Illumina’s challenge.
...
A webcast presentation today for customers of the company will likely put many of those fears to rest. Oxford Nanopore’s Chief Technology Officer Clive Brown announced an upgrade to its devices—a new pore the company has been describing in presentations as R9. Brown revealed that R9 is CsgG, a membrane protein derived from Escherichia coli, but present in many bacterial species.



http://www.sciencemag.org/news/2016/...new-technology

[ymc]

If the accuracy of R9 chemistry plus RNN algorithm is true, then both PacBio and Illumina will be dead.

However, judging from ONT's past record, this might well be achievable but the timeline is more likely to be one or two years in the future. As a result, this lawsuit can still be a distraction for ONT for the time being IMHO.

[luc]

If the accuracy of R9 chemistry plus RNN algorithm is true, then both PacBio and Illumina will be dead.

I believe the track record of ONP/Clive Brown announcements is pretty abysmal.

[ymc]

I believe the track record of ONP/Clive Brown announcements is pretty abysmal.

Yeah, so far they only delivered Minion and it was at least two years late.

[AllSeq]

If the accuracy of R9 chemistry plus RNN algorithm is true, then both PacBio and Illumina will be dead.

Matching the error rate of PacBio would indeed be pretty impressive, but that alone is insufficient to kill off PacBio or Illumina. The question is does the ONT error profile match that of PacBio? In other words, is it stochastic and relatively free of GC bias so that at sufficient coverage the consensus error rate drops to very low levels (like PacBio). Or does it suffer from systematic errors that create a ceiling to the consensus read quality? Have these questions been definitely answered yet?