Subject: In directed evolution, how are sequences of interest identified?

Hello,

I've been reading much on the two general protein design techniques: directed 
enzyme evolution and the rational evolutionary design route. However, I'm 
having a tough time understanding the former concept. My question is, in a 
directed evolution experiment, is the entire protein of interest mutated 
randomly or must the scientist identify a specific sequence of interest (or 
region in the protein) that they wish to modify/mutate?  If the latter is 
true, how does one come about identifying the specific sequence (not a priori) 
in a protein with literally hundreds to thousands of bps?

Your time and help is deeply appreciated.
-Amy

Re: In directed evolution, how are sequences of interest identified?