I’m having a problem producing a good idealized PDB file for a monomer with about 700 residues. The resulting idealized structures have positive scores that are two orders of magnitude larger than the wildtype and they look somewhat unraveled.
Is the number of residues a factor in causing these problems?
I also noticed that the wildtype protein had positive score above 800. Does this indicate that I should minimize the starting pdb structure before I idealize it or is that not a factor?
I don’t think you should have to minimize first. What command line are you using to idealize? If the starting pdb has very bad geometry it may cause a problem. Is the starting structure from teh RCSB?
there’s two idealize protocols in Rosetta++, one written using pose and one written before pose. In my experience the pose-idealize works much better. I believe you simply have to add ‘-pose’ along with ‘-idealize’.
I wouldn’t minimize the starting structure before idealizing. Keep in mind that, barring chain-breaks, idealizing is much more difficult at larger chain lengths and 700 is pretty large.
also for large multi-domain proteins, its sometimes better to idealize the domains individually and then splice the idealized structures together afterwards.
.
