Big numbers, new tools at Keystone
One thing's certain at the Keystone Symposia on linkurl:Structural Genomics;http://www.keystonesymposia.org/Meetings/ViewMeetings.cfm?MeetingID=817 and linkurl:Frontiers in Structural Biology;http://www.keystonesymposia.org/Meetings/ViewMeetings.cfm?
One thing's certain at the Keystone Symposia on linkurl:Structural Genomics;http://www.keystonesymposia.org/Meetings/ViewMeetings.cfm?MeetingID=817 and linkurl:Frontiers in Structural Biology;http://www.keystonesymposia.org/Meetings/ViewMeetings.cfm?MeetingID=816 running this week in Keystone, Colorado: there's some seriously big science going on in the world of structural biology.
Aled Edwards of the University of Toronto rattled off the program goals of the linkurl:Structural Genomics Consortium;http://www.sgc.utoronto.ca/ (SGC), of which Edwards is the chief executive: 50 structures solved between July 2004 and June 2005; 112 structures in '05-'06; and 224 in '06-'07. Does that sound laughable? Hardly.
With facilites at the universities of Toronto and Oxford and the Karolinska Institute, the consortium has since its inception solved some 180 structures at an average $125,000 per. The linkurl:RIKEN Genomic Sciences Center;http://www.gsc.riken.go.jp/eng/group/protein/index.html is putting up even bigger numbers. According to project director Shigeyuki Yokoyama, the Center, with some 40 NMRs (including 3 900-Mhz instruments and 14 800s), solved 1,347 structures between April 2002 and October 2005, 700 of them by NMR.
Impressive as those numbers are, I was most interested in a throwaway comment Yokoyama made near the end of his talk. In a session on protein expression and purification, Yokoyama was describing how an E. coli cell-free protein expression system has streamlined structural biology efforts at RIKEN. Then he mentioned that a protein extract derived from Drosophila S2 cells could be used for in vitro RNA interference.
Imagine how useful something like that could be to the RNAi community. Conventional wisdom holds that you should make several siRNAs per gene, as not all of them will work. This system could speed up this trial-and-error process: Rather than optimizing conditions in cell culture, researchers could test their siRNAs in vitro to see if they block translation first. RIKEN is filing patents on the technology, Yokoyama told me later, so commercialization could be forthcoming.
Brian Marsden, of the Structural Genomics Consortium at Oxford, presented linkurl:another nifty tool;http://www.sgc.ox.ac.uk/iSee/ during the post-seminar poster session. Recognizing "the need for integrated data visualization," Marsden's team has developed iSee, a sort of knowledge base-cum-Wiki for structural genomics.
Right now, biologists hoping to leverage structural biology data must integrate information from disparate sources: papers in Adobe Acrobat, structures in RasMol, sequence annotations in their web browser, and so on. iSee integrates all this data into a single data file and browser.
Experts annotate the structure with methods, key points of interest, alignments, and so on. Users can then browse those structures, suggest new annotations, even dispute existing ones. It's not quite a Wiki, but it's close (albeit slower). The files are small enough to be emailed, and the browser runs under Windows, Mac OS X, and Linux.
Marsden says he envisions iSee (or something like it) as a new paradigm in structure publishing. There's at least one problem with that, however: The program needed to create the files isn't free. According to Marsden, it'll set you back about $100.