On Fri, Nov 19, 2010 at 11:33 AM, Jacob Keller < j-kell...@fsm.northwestern.edu> wrote:
> I took a look at his slides--looks pretty interesting, and I would > have liked to have heard the talk! I am not sure I would be ready to > advocate having no more journals, but I am interested in thinking > about the idea. Imagine just putting your data up on the website, and > you're finished! And then wait for comments/links to show up, I > guess--would that be the metric for grants/funding? I also wonder what > could be done so that the record would not be changed? > You can't rely on personal academic websites for long-term archival of anything. This is one reason why institutions such as the PDB and GenBank (among others) exist, to ensure that published data is accessible at a central location, and there is considerable economy of scale from doing it this way. What's missing, of course, is any associated content like you propose, but organizations like the structural genomics groups (at least in the US) are already halfway there, since they immediately release their structures, presumably before they even write a paper. Large genomics centers do the same with their raw sequencing data. However, structural genomics is a very different enterprise from traditional hypothesis-driven science, and I'm not convinced that the immediate-release model will ever be applicable to most academics. Anyone with sufficient technical skill and proper resources can generate raw data; figuring out what it means, and what follow-up experiments need to be done, is much harder. There are many ways to incorporate collaborative annotation and other unconventional methods of publication; I think PLoS ONE (which allows comments) is an excellent step in the right direction. However, until we start seeing ribosome or membrane-protein structures published this way, everyone with an eye on a future academic career is going to prefer the traditional route. -Nat