Well, I'm not at AGBT. The sobering fact is that even if I had registered, I wouldn't be at AGBT -- Logan Airport closed yesterday morning and won't open until sometime tomorrow. Even with a plow doing the driveway, there was plenty of digging to be done here.
There wasn't a lot going on with Twitter today about the conference, but a few nuggets of interest. Actually, the first order of business seemed to be a debate on whether to use the hashtag #agbt or #agbt2011. While the shorter version seems to dominate, a few folks are using the #agbt2011 alone and then there are mugwumps tweeting both. I haven't started using my Perl client yet, but this is one more advantage to having it -- I can easily merge both streams (which I can probably do with the standard interface, but sometimes it's easier to write one's own code than read someone else's documentation!)
A lot of the news out of AGBT today came from vendors -- various new announcements. For example, RainDance announced a new product for targeting FFPE DNA and is partnered with NuGen for upstream DNA amplification. LifeTechnologies is giving away another Ion Torrent PGM sequencer with a conference "Triathalon" -- though my non-attendance is irrelevant since it's open only to academics (ahem!). LifeTech also announced another "Grand Challenge" contest, $1M to someone who can sequence an entire human genome AND transcriptome from a single cancer cell using SOLiD.
Talks tonight were from some of the major sequencing centers. Harold Swerdlow from the Sanger apparently covered both their current Illumina operations as well as trials of the Pacific Biosciences instrument. The Sanger is now generating a jaw-dropping megabase of data per second; when I was a grad student one of the grant proposals was to build a megabase per day facility. Or, if you'd prefer, apparently that works out to two terabases per week. Yow!
On the Pacific Biosciences side, they are apparently getting 28 million mappable bases from reads averaging 1650 nucleotides, implying around 17K reads per 40 minute run. Definitely a lot of things one could do with that, but also it won't really compete in a lot of arenas -- particularly with only 86% accuracy. Still, long noisy reads could be very useful for scaffolding smaller ones and with 500 base fragments this would mean three passes in circular consensus mode.
A few other things that caught my eye. Harvard Biopolymers uses Twitter to let people know of spare Illumina lanes. Galaxy seems to be quite a hit and has a thriving ecosystem of add ons -- and easy install on the Amazon cloud. The Broad showed even more interesting cancer data -- though no details tweeted. Actually, their multiple myeloma sequencing paper has been allegedly in press since a conference I attended last June -- will it ever see daylight?
Tomorrow should be meatier -- talks all day. Since I won't be there, I also won't be faced with the horrid dilemma posed by the concurrent sessions in the evening. One is on all sorts of blue sky sequencing technologies and might make the pulse race. The other has lots of talks on using existing technologies, first with a string of cancer talks and then a bunch of RNA-Seq. Friday has a similar split -- one session on techniques and one on applications (including 1 of two dog talks in the confernece). Still, I'd love to struggling to decide!
Thanks to everyone who has tweeted! If only there was a way to get the slides too. Some meeting organizers might worry that too much off-site access to a meeting might hurt business. But with Marco Island, it's hard to see much worry. First, it's a cool locale. Second, watching remotely isn't the same as being there. Third, we can't see or touch any hardware (both a PGM and a MiSeq are apparently onsite). Finally, those of us remoting in can't network very effectively -- and that is the most fun of such conferences.