ramachandran_restraints
Dear all, I am refining a structure to 2.9 A, with pretty bad experimental data (P1, ~5400 residues in the A.U. (4 mol), anisotropic, fairly incomplete...). I have been running phenix.refine with NCS, 1 ADP per residue, lower target weights, TLS. My R/Rfree are stuck at about 2.4/3.05. I was getting around 2% ramachandran outliers at every cycle so I have been giving the nightly builds a shot. Now I get 0.2% outliers without changing any parameters for the ramachandra_restraints but it is forcing some of the peptide bonds out of the density. What can I tweak to improve the refinement? Thanks all in advance! Jose José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal "It's very hard to make predictions... especially about the future" - Niels Bohr
Hi Jose, Is it possible that the density is inaccurate enough that the "forced out of density" peptide bonds might actually be closer to the correct positions than where they were previously? At resolutions in this range if the overall ADP's are high and the map is not that clear sometimes moving things out of density but into geometrically likely configurations could possibly make the model better. All the best, Tom T
Dear all, I am refining a structure to 2.9 A, with pretty bad experimental data (P1, ~5400 residues in the A.U. (4 mol), anisotropic, fairly incomplete...). I have been running phenix.refine with NCS, 1 ADP per residue, lower target weights, TLS. My R/Rfree are stuck at about 2.4/3.05. I was getting around 2% ramachandran outliers at every cycle so I have been giving the nightly builds a shot. Now I get 0.2% outliers without changing any parameters for the ramachandra_restraints but it is forcing some of the peptide bonds out of the density. What can I tweak to improve the refinement? Thanks all in advance!
Jose
José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal
"It's very hard to make predictions... especially about the future" - Niels Bohr
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Hi Jose, what happens if after a round of refinement with Ramachandran restraints you do another round of refinement without Ramachandran restraints? Pavel. On 12/28/10 4:56 AM, José Trincão wrote:
Dear all, I am refining a structure to 2.9 A, with pretty bad experimental data (P1, ~5400 residues in the A.U. (4 mol), anisotropic, fairly incomplete...). I have been running phenix.refine with NCS, 1 ADP per residue, lower target weights, TLS. My R/Rfree are stuck at about 2.4/3.05. I was getting around 2% ramachandran outliers at every cycle so I have been giving the nightly builds a shot. Now I get 0.2% outliers without changing any parameters for the ramachandra_restraints but it is forcing some of the peptide bonds out of the density. What can I tweak to improve the refinement? Thanks all in advance!
Jose
José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal
participants (3)
-
José Trincão -
Pavel Afonine -
Thomas C. Terwilliger