A machine that speeds up evolution is revolutionizing genome design
wyss.harvard.edu
wyss.harvard.edu
This concept, like so many other techniques developed by that lab, is very impressive. The last major contribution I was on the cutting edge of (back in 2008) was the ability to predict and design proteins with totally different amino acid sequences that performed the same function biologically, which was revolutionary as it showed not only that you can replace almost every amino acid in a protein with a different residue and be biologically equivalent. It also proved that it was not the actual amino acid sequence that performed the function but the type of residue (hydrophobic, hydrophillic, acidic, basic, large or small). Of course, you can't overcome a well placed disulphide bond but the ability to do this basically nullified any possibility that a company or organisation could patent a particular protein, as you could simply replace almost the entire sequence of that protein and still have the same functionality.
I'm sure I have missed numerous other advances over the past 7 years as well.
Some cool stuff coming out of that lab.
Edit: it looks like this was from 2009 so not as cutting edge as I thought
They seem to be doing well at improving luminosity through directed evolution, as can be seen in the graph in the December [1] update. According to their last update[2], they've increased the automation since then.
[1] http://blog.glowingplant.com/post/105558826472/happy-holiday...
[2] http://blog.glowingplant.com/post/118481662132/glowing-plant...
Viruses are well ahead of us in that.