Despite having good fold-recognition templates, we had a very hard
time getting a complete model for this protein.  We were also unable
to repack the sidechains with Rosetta, since the version we had
installed did not have MAX_RES set high enough to handle this large a
protein, and we did not have the time to acquire and compile a new
version. 


One of the earlier versions of our model (try10-opt2) also scored well
with an unconstrained cost function.

We are submitting it as our model 2.