Saturday, February 21, 2015

#AGBT2015 Preview

The annual genomics party on Gulf of Mexico beaches named AGBT runs next week, and already there have been some speculations flying.  I'd better dash something off before I'm any later to the preshow -- or more importantly before I get contaminated with embargoed information.
The one piece of utterly solid information in my hands is the personal side: after two straight years of actually attending, I will be watching this one via participants' blogs and tweets. I could give a long song-and-dance about my complicated feelings towards AGBT, but the reality is that the short registration window tripped me up.  I tried registering when the website was supposed to be live -- but it wasn't -- and then had all my off-work attention focused on preparing for a family reunion where we were pre-celebrating the ninetieth birthday of the family matriarch.  My main goal this year is to not repeat my skiing performance the last time I didn't go; I carry about a dozen pieces of metal around with me constantly thanks to that little oops.

A blog called DeciBio has released one set of rumors, which can be charitably characterized as an ill-refined blunderbuss approach. I'll smack down a few of these below and then stick my own neck out in a few places.

The one hit would be 10X Genomics, which is a Bronze Sponsor and has made clear that they will be further unveiling their synthetic long read technology which they announced at the J.P. Morgan conference in early January.  10X had some less fun headlines recently when RainDance slapped them with a patent lawsuit over microfluidics, which is surprising timing as such lawsuits are more typically launched when a product is actually hitting the market and there is a revenue stream to be latched onto.  

Pacific Biosciences splurging for a Gold sponsorship has engendered rumors in that DeciBio piece as well as elsewhere (e.g. stock analyst Amanda Murphy) that PacBio will have a significant announcement.  That would certainly be exciting, but if they have one the security has been excellent.  Other than the persistent rumor of a desktop-sized instrument, there hasn't been chatter, and my guess is such a baby RS is so technically challenging it is unlikely to be launched (PACB: please prove me wrong!).  Also, public companies seem to be weighting their big announcements more to JP Morgan than AGBT.

For that last reason, I doubt Illumina (Silver sponsor) will make any huge announcements; JPM has been their forum for a number of years now. There will be details on their new boxes and probably more notes on the NeoPrep library system.  

The DeciBio piece throws out the possibility of big announcements from a long list of unlikely candidates, since these companies have neither sponsorships nor talk slots.  So I think it unlikely BGI will launch their lightly announced sequencing boxes or NabSys will roll out a new platform or chemistry or Clic will unveil a smaller library prep box.  QIAGEN will probably not have any update on when their sequencing box will actually see light. We can certainly rule out announcements from Oxford Nanopore; ONT has made it clear that they will save big splashes for their upcoming London Calling meeting. There are at least three talks by MinION advance access participants on the agenda, which is the way ONT prefers to have their horn tooted at these meetings.   I don't see any exciting new platform companies on the agenda, so if any interesting new techs are described at AGBT, it will have to be in the poster session (whose abstracts don't seem to be up) -- my guess is there are a lot of gems hiding on the posters, which is often the most productive area for me.

Roche is a Bronze sponsor, so they might make announcements in their suite.  On the other hand, that could be just habit on their part plus while they've shuttered the 454 platform they do have exome capture products and such to continue to market.  I'd love for Roche to release information about their Genia acquisition, but am not optimistic.

One of the more bizarre quotes from an unnamed survey respondent in the DeciBio piece concerns ThermoFisher and the Ion Torrent platform, from someone who apparently missed the announcement that the PIII chip is dead.  Same quotee called 400bp reads "long reads", which is itself problematic. Given PacBio and now Oxford Nanopore results, long is at minimum 5K in my book; with a full-length 48K lambda read now enshrined in the peer-reviewed literature, 5K is no longer so impressive.  I'm guessing the Ion team will mostly continue to spread the word that their HiQ chemistry is a reason to look again at Ion Torrent, a hard sell given the harsh skepticism of hard-core Illuminati.  It would be nice to see Ion putting more energies on their isothermal template prep chemistry, but not holding my breath there. I'll try to post a longer pondering on Ion's position in the near future.

One possible emerging trend: both Fluidigm and Nanostring have been making noise that they will be describing systems for simultaneously assessing both protein and nucleic acid markers in the same sample (same single cells in the case of Fluidigm).  There's also a talk on simultaneous sequencing of RNA and genomic DNA or RNA and methylome.  A droplet-based single cell sequencing system ("DropSeq", naturally) is also on the agenda.

Looking over the agenda, it is as dominated (or perhaps more so) by genomic biology as the past, perhaps is simply a sign of the maturity of the field.  That's one of my complexities with going: as a scientist and blogger it's great fun and very stimulating, but when the day job is funding such a junket it's a bit frustrating to have so little to bring home of relevance.  There is actually a natural products talk, but it's by company adviser Michael Fischbach so I'm quite familiar with it  (he's a good speaker; don't skip it!).   There's at least two other microbiome talks on the agenda as well (I remember suggesting more microbiome in the survey form last year; nice to see this happen). I also wish I could hear about a 70X coverage goat genome on PacBio or epigenetic responses to spaceflight.

As far as tech talks, in addition to the two MinION talks I mentioned above, there's a talk from the Broad on rapid, full-length Ebola and Lassa sequencing -- it will be interesting to see if that is yet another MinION talk or simply a fast workflow for other platforms.  A number of talks on single cell analysis, which is really starting to be a big thing (though very pricey). Promega has a talk with "automated sample preparation" in the title, which could be interesting.  On the software side, Richard Gene Myers is speaking further on "Low Coverage, Correction-Free Assembly for Long Reads" which I wish I could attend.  There's also a talk on a graph-based aligner that would be cool to hear.

So, an advance thank-you to all the attendees who will blog and tweet.  Enjoy the hospitality suites and the talks, and make sure to enjoy the natural beauty of Marco Island.  And if any companies want to make me look silly by making a big splash, go for it!


Dale Yuzuki said...

Many thanks for clarifying the Decibio post - I thought it was a bit odd too, but they do what they can. In general though they do very good work - Thermo Fisher has purchased several of their reports and I've found little to disagree with (and much to learn).

Will be missing your live Twitter commentary this year! I hope you haven't gotten too depressed from all that Boston snow. :)

BTW I posted this last week about #AGBT15:


Mike the Mad Biologist said...

Maybe PacBio will be displaying its supposed increase in output/plate?

Cliff Beall said...

"Illuminati", good one!

Verdant Force said...

Informative as always Keith. Poster abstracts are up on guidebook:

PharmacoGenomics said...

Great post. As someone who has only fairly recently learned about precision medicine/genetic sequencing/ pharmacogenetics it's blogs like yours that help me learn about this stuff.

Seriously I appreciate the effort you must put into making this blog and filling it with pertinent information.

Wish I could've attended #AGBT2015 as I'm sure it's fascinating. Oh well, I guess there's always 2016.