This year’s ASHG conference turned out to be a treasure trove of updates from the major next generation sequencing platforms. (So much so that we have to wonder if there will be any surprises at AGBT in February…) We didn’t attend this year, but thanks to the many, many tweeple who attended, we were able to follow along pretty well. In this blog post I’ll focus on the new announcements and what they mean for people with sequencing projects, but if you’d like to read some more comprehensive blog reviews, I’ve listed some good ones at the end.
The most excitement seemed to center around Oxford Nanopore (ONT). While they didn’t show any sequencing data, update the launch specs or give a solid launch timetable, they did bring toys to the party! A bunch of lucky attendees got to play with some MinIONs, load them up with DNA and watch squiggles appear magically on the screen. (Participants were told to trust that the squiggles actually equated to DNA sequence, but no actual sequence data was returned).
In lieu of a launch date, ONT instead announced an early access program. For a refundable $1000 you get a MinION and ‘some’ flow cells. First you have to run test samples provided by ONT to make sure that the system is running properly. Once it passes the test, you can run your own samples. (I’ve heard some people complaining about the need to run test samples first, but that’s actually pretty standard stuff for a beta test – it’s really useful to see results from the exact same samples across multiple test sites.) After the initial supplies run out, they’ll let the testers purchase additional flow cells at $1000 each.
And while ONT hasn’t shared any of their own data, they’re letting the testers publish their own data as they see fit (after successfully passing the testing period, of course). That means we should start seeing some real data in the next few months. We can’t wait! Best guess is that it will be pretty messy at first, but hopefully all of the testing will lead to improvements. Perhaps the most useful thing will be getting some long read data out ‘into the wild’ so some clever folks can start working on the new algorithms that will surely be needed to handle it.
Here are some other thoughts on ONT’s announcements:
Yaniv Erlich does a really nice job of explaining the ONT sample prep at his Future Continuous blog.
Lex Nederbragt compares the MinION to a high tech washing machine he wouldn’t buy over at In Between Lines of Code.
Keith Robison compares the MinION to a ‘build your own’ Sinclair ZX-81 computer over at Omics! Omics!
Ion Torrent also decided to make a big splash at ASHG. In fact, they took the early lead during their Ion World pre-conference. They seemed to spend a lot of time explaining how hard it is to make the PII chip. Like, really hard. They’re making progress, but there’s been a delay. No, not that one, I mean a new delay. Originally slated for the end of 2012, then March 2013, then mid-2013, then late 2013, it’s now slated for May 2014. And the launch specs will be 20-30Gb per run instead of the original 100Gb per run (which was going to give us that magical $1000 genome in a single day). It kind of makes you wonder if the PIII will ever see the light of day. So, with nothing to launch today, Ion Torrent figured they had to give their customers something to keep them occupied until May. So they launched the “Technology Access Program”. This is sort of like ONT’s beta access program, except any Ion Torrent customer can participate. So what do customers get access to? Not the PII – it doesn’t really exist yet. Instead, they get to buy unsupported versions of “Ion Isothermal Amplification Chemistry” (what we used to know as ‘Avalanche’) and “Hi-Q Sequencing Chemistry”. Hi-Q is really the big news. It’s a new enzyme that they say reduces the homopolymer errors by 90%. It’s available in November in the Technology Access Program with an expected launch date in July 2014 (no similar date supplied for Avalanche). Now the bad news. It’s only for the PGM. It turns out the PGM and Proton have been using different enzymes, and they started the Hi-Q improvements on the PGM as that’s the Ion Torrent platform that’s more closely associated with the clinical space (and where homopolymer errors are much more damaging). There’s no reason to think they won’t bring the same improvements to the Proton enzyme, but there isn’t a target date for that yet. Dale Yuzuki (from Life Technologies) covered the Ion Torrent announcements over at his blog.
That’s it for part 1. Next time we’ll cover announcements from Illumina, PacBio and BioNano Genomics. In the meantime, here are some blogs with more comprehensive coverage of the conference:
Yaniv Erlich’s Future Continuous
Dan Koboldt’s Mass Genomics (edit – link no longer working)
Brigitte Ganter’s enlightenbio blog
Aaron Krol at Bio-IT World