I was using 1.6.1. Now I am trying 1.7.3 now to see if it will fix my problem. I actually used ramachandra restrain in coot during building those loops, but some residues still changed back. That's why I was considering for this option. Thanks. Nian On Tue, Apr 3, 2012 at 3:47 PM, Nathaniel Echols wrote:

On Tue, Apr 3, 2012 at 1:38 PM, Nian Huang wrote:

...

Is it possible to limit Ramachandran restrain to specified residues in phenix.refine? I have a couple of loops with very poor density. The refinement process always made some residues fall out of ramachandran plot no matter how I tightened the weight and adjusted the residues back. Because the diffraction is about 2.3 A, I don't want the ramachandran restrain to impact the rest of structure. Thanks.

Which version of Phenix are you using?  If it's 1.7.1 or earlier, you need to update to the latest version ASAP, because this will probably fix the problem.

If you're already using the latest version, the parameter rama_selection will do what you want.  But you really shouldn't need Ramachandran restraints at 2.3A resolution - and I'd advise against using them entirely, at any resolution.  They're just too dangerous and lead to unnatural structures.  You also probably don't want to be running real-space refinement for this structure.  However, when you're rebuilding, using Coot's Ramachandran restraints together with its local real-space refinement should be sufficient to give you a good starting model, which Phenix will hopefully not screw up.

-Nat _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb