The whole point of this is to arrive at the "Holy Grail" for biochemistry/medicine. The ability to predict structure from sequence data alone is what we're after. It's easy to get a protein sequence for a molecule of interest, but to get the structure (so you can BEGIN to think about function) takes a long time and lots of luck using techniques like X-ray crystallography and NMR.
What do I win?
Potentially a lot!