Hi Dr Afonine, Thank you so much for the reply! I also posted on Phenix mailing list and Dr Nathaniel Echols suggested me to turn on X-ray/stereochemistry weight optimization (for non-GUI users, optimize_xyz_weight=True). So, I ran the refinement again with default settings + Rigid body+ Simulated annealing (Cartesian) + Optimize X-ray/stereochemistry weight. And the statistics improved! Also, the gap between R-free and R-work becomes smaller, less than 6%. See below: R-work=0.2348, R-free=0.2855, Bonds=0.003, Angles=0.919. Ramachandran outliers: 2.5%, Ramachandran favored: 88.1%. Rotamer outliers: 10%, c-beta outliers: 0 clashscore: 9.06, overall score: 2.84
Now, I am wondering whether there is any suggestion about what to try and check next to further refine the structure. Thank you so much! Best, Wei On Sun, Jul 28, 2013 at 3:25 PM, Pavel Afonine <pafon...@gmail.com> wrote: > Hello Wei, > > I have a refined structure with phenix. >> > > FYI and as a reminder: there is Phenix mailing list for Phenix-specific > questions. > > >> The resolution=3.2 Å. After refinement, the R-work=0.2186, R-free=0.2864, >> Bonds=0.010, Angles=1.515. >> > > Large Rfree-Rwork gap indicates overfitting, which in turn makes me think > that refinement strategy you use is not optimal for the data and model you > have. If you send me files off list I may be able to suggest a better > option. > > >> Rotamer outliers: 14.5%, c-beta outliers: 2 >> > > This suggests me that you may not be using a recent version of the > program, since recent versions fix problems like this in real space and > apply rotamer-specific and cbeta restraints during refinement, and these > restraints are updated dynamically. > > >> To fix the Ramanchandran outliers, in coot, by changing the phi-psi of >> the outliers, I could manage to move most of them to the allowed positions. >> But, after another refinement in phenix, those outliers come back. >> > > This is a scenario when you may want to apply Ramachandran restraints > locally to residues that are notorious outliers. Steps to do this are: > - first, fix Ramachandran outliers manually, > - then, running subsequent refinement include these residues into > Ramachandran restraints. > > However, I think you may even not need to do this if you use optimal > refinement strategy (see above): the fact that you get notorious outliers > may be a result of not using correct refinement strategy. > > Good luck! > Pavel >
