data ecologies

almost three years ago now, i was able to get ed wilson, harold varmus, and jeff sachs to google to make a pitch for what we then called biosearch--a facility for biological information that would enable data from multiple fields to interact. not much came of that, though ed was one of the TED prizewinners for 2007, and his TED wish was the encyclopedia of life, now being built by the EOL consortium. this was our general architectural representation of biosearch:

i think it's a good general framework for thinking about what biological data management should look like. thoughts?