Friday, December 02, 2016

Oxford Nanopore New York City Meeting, Day 2

The second and final day of Oxford Nanopore's New York User Meeting ran today.  I've again been mining tweets, since I wasn't on site.  Oxford itself has posted a summary of Day 1, which has the enormous benefit of the author being present! I'll make a few quick summaries.  The tweets for today can be found in two semi-thematic Storify entries: one gives general coverage and of ONT's demos, whereas the other covers ONT's technical talks and talks by users.

User talks at the meeting continued the pattern of hitting a wide range of MinION's strengths: portable sequencing, rapid sequencing, long reads and RNA.  While in terms of market share MinION is still a niche player, it's occupying a very large number of niches!  Also the issue of regulatory approval for diagnostic applications, which I had fretted wasn't receiving enough attention, is going forward (at least in Europe).

A new addition to MinION lore is Jane Carlton sequencing while traveling by taxi in India! I had once suggested to Ion Torrent that they generate an ad loosely based on Indiana Jones: intrepid hero escapes attackers, jumps in Land Rover and sequences while bouncing over rough terrain. Ion never took me up on it; perhaps ONT can take it up.  Reminds me of the classic ad for the AMC Pacer, though with a MinION an MG would be more than enough.

Andy Heron from ONT discussed further the Cas9 capture system which Clive Brown had introduced in the Wafer Thin Update.  The protocol is interposed within the new five minute rapid 1D prep. DNA is fragmented and primed for adapters, then incubated with Cas9 for 20 minutes.  An SPRI cleanup (~20 minutes) and a bead binding step (~20 minutes) and then the final adapter-adding step (~5 minutes).  So the promised wallclock time is on the order of 1:10.  Input DNA is apparently on the order of 250 nanograms.  Arwyn Edwards tweeted that this should enable performing 16S profiling on a mixed bacterial population, but without any PCR.

Or you could sequence the rRNA directly: Andrew Smith from UCSC showed direct sequencing of E.coli ribosomal RNA.    Currently the protocol requires an hour, but it is believed that can be halved.

MIT's Julie Hachey explored the possibility of putting a MinION-based system on a future Mars rover. She showed success sequencing DNA containing inosine, which has apparently been seen in DNA from meteorites.  It's an interesting idea, but certainly nanopore reagents will be much more sensitive to a long, cold bath in space than typical NASA instrumentation!

Sophie Zaaijer showed how she could identify her adviser Yaniv Erlich in the first 14 minutes of sequencing, using 115 SNPs searched against a database of 31,000 DNA profiles.  Fears of waiters surreptitiously swiping your credit card with a scanner in their pocket can now be replaced with the terror of being identified from your salad fork before you've finished dessert!

Large genomes besides human?  Christian Henkel described assembling the European eel; Ioannis Ragousis the olive tree fruit fly.  If the new Scrappie basecaller can really solve the homopolymer problem, then MinION may become a real force for de novo sequencing of all but the biggest genomes.  Even if users can expect only 4 gigabases per flowcell (versus the 10 which ONT achieves; their numbers have not been reliably reproduced in the field), that would enable 40-50X coverage of many fungal and small metazoan genomes on a single flowcell.  Of course, one can always blitz through multiple flowcells, as the human sequencing group did

I'll apologize again for having less stamina then required to do justice to the talks I haven't covered ; lots more interesting stuff that made me wish again I could have attended.

The third London Calling meeting will be the 4th and 5th of May.  If ONT can deliver VolTRAX (units are showing up), direct RNA sequencing kits (promised before end of 2016) , the Scrappie basecaller and low-cost Flongle mini-flowcells ("soon") and PromethION flowcells ("shipping soon") then that should be a very exciting meeting indeed.


Anonymous said...

Getting a bit exhausted with them. So much announcements one after another without giving time for applications to percolate and establish. Starting to look amateurish.

Ian said...

"DNA containing inosine, which has apparently been seen in DNA from meteorites."

Unless I missed some really big news, DNA has not been found in meteorites. It's nonstandard bases (inosine) that has been found in meteorites.

Keith Robison said...

Ian: You didn't miss anything; I wrote badly.

Geneticist from the East said...

I have a question regarding Andrew M. Smith et al from UCSC claiming direct RNA sequencing. If it is true, does that mean either they or ONT recalibrated the basecaller to call A,C,G,Us?

It will be great if someone can post some screenshoots of the slides of his talk.

Keith Robison said...

Geneticist: Best reference for the direct RNA sequencing is the BioXiv preprint for the method. "The basecaller used in this work is a naive HMM based solely on event means and has not
been extensively optimised for the purpose of basecalling RNA data. The MinION data is far richer
than this simple model supports. "

Andrew Smith said...

Hello Keith and Geneticist from the East,

All talks from the conference, including mine, were filmed. They should be available through the ONT website in the near future.

Andrew Smith

Geneticist from the East said...

Thanks Keith and Dr Smith for the replies.

Now I understand that only motor protein is changed but the flowcell remains the same. Of course, the basecaller also need to be changed.

Dr Smith, does ONT now also have an RNN basecaller for RNA?