[phenixbb] High B-factors and High Clash Score

[Edwin Lazo]

Hi all,

I am refining a structure that diffracted to 3.5 Angstroms and contains 
6 monomers in the ASU. It is noteworthy to mention, in certain monomers, 
half their residues are missing (in other words no density). Initial 
b-factors after molrep were 118.585, R=0.32, and Rfree=0.40. Initial 
refinement was done with refmac and:

mean b-value=117.344
R=0.31
Rfree=0.40

Refinement with Phenix 
(phenix-installer-dev-2054-intel-linux-2.6-x86_64-centos5) was performed 
and the Ramachandran plot was bad:

OUTLIERS : 6.52  %
ALLOWED  : 13.67 %
FAVORED  : 79.82 %
ROTAMER OUTLIERS : 7.18 %
CBETA DEVIATIONS : 8

And final R=0.38 and Rfree=0.41.

I then improved Ramachandran plot in coot and follow Pavel's suggestion:

/  - add H atoms:
/>>/
/>>/  phenix.reduce model.pdb > model_H.pdb
/>>/
/>>/  - then run a quick geometry regularization:*(**T**his part I**did in coot. phenix.**pdb**tools model_H.pdb --geom and phenix.pdbt**ools model_H.pdb --geometry**-**restraints do not work)*
/>>/
/>>/  phenix.pdbtools model_H.pdb --geom
/>>/
/>>/  - then strip off H:
/>>/
/>>/  phenix.reduce -trim model_H_regularized.pdb > start.pdb
/>>/
/>>/  - then run refinement using NCS and Ramachandran plot restraints:
/>>/
/>>/  phenix.refine start.pdb def_data.mtz main.ncs=true --overwrite
/>>/  output.prefix=def2 main.number_of_mac=5 ramachandran_restraints=true
/>>/  rama_potential=emsley strategy=individual_sites+**individual_adp
/>>/
/>>/  - then take outcome of previous refinement and run the same refinement
/>>/  plus weights optimization:
/>>/
/>>/  phenix.refine def2_001.pdb def_data.mtz main.ncs=true --overwrite
/>>/  output.prefix=def3 main.number_of_mac=5 ramachandran_restraints=true
/>>/  rama_potential=emsley strategy=individual_sites+**individual_adp
/>>/  optimize_xyz=true nproc=6

/My current stats are as follows:

Final R=0.35
Final Rfree=0.38
ALL-ATOM CLASHSCORE : 173.74
RAMACHANDRAN PLOT:
OUTLIERS : 2.27  %
ALLOWED  : 4.61  %
FAVORED  : 93.13 %
ROTAMER OUTLIERS : 36.26 %
CBETA DEVIATIONS : 6
Wilson B-Value: 108.4
Mean B-value: 108.997

I am unsure how to proceed from here. I am planning on removing some side chains which have no density and then refining with Rama. restraints.

I can also send my mtz and pdb if that would help.

The stats when I processed the data are as follows from xia2 -3d (xds, xscale):

High resolution limit                   	  3.50	 15.65	  3.50
Low resolution limit                    	109.96	109.96	  3.59
Completeness                            	 94.1	 90.5	 91.5
Multiplicity                            	  4.8	  5.1	  3.2
I/sigma                                 	 11.3	 26.5	  2.4
Rmerge                                  	0.118	0.068	0.458
Rmeas(I)                                	0.144	0.080	0.605
Rmeas(I+/-)                             	0.144	0.083	0.584
Rpim(I)                                 	0.062	0.036	0.311
Rpim(I+/-)                              	0.081	0.047	0.358
CC half                                 	0.992	0.994	0.791
Wilson B factor                         	90.052
Anomalous completeness                  	 86.9	 93.6	 71.3
Anomalous multiplicity                  	  2.4	  3.0	  1.8
Anomalous correlation                   	-0.262	-0.440	-0.052
Anomalous slope                         	1.005	0.000	0.000
Total observations                      	169836	2234	7765
Total unique                            	35071	442	2464
Assuming spacegroup: P 21 21 21
Unit cell:
73.010 180.780 219.910
90.000  90.000  90.000

Any help would greatly be appreciated,

-- 
Edwin Lazo
Scientific Associate
Life Science and Biomedical Technology Research Resource
National Synchrotron Light Source II Building 745
Upton, NY 11973
Office:(631)344-7247
Fax:(631)344-5059

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20150724/a6fa8372/attachment.htm>