*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Dear list members,
I am currently refining a molecular replacement solution (3 mols/AU)
against perfect merohedral twinned data at 2.45 A resolution. After one
round of rigid body refinement, rebuilding and ten cycles of CGLS
refinement without the BLOC 1 statement, the maps look very nice and I
can see clear difference density for an (expected) ligand (R/R(Free)
0.21/0.27). However, the ramachandran plot shows a large number of outliers.
Currently, only about 85% of the residues are within the allowed regions.
As I have started from a pretty complete and stereochemically sound mr-solution,
I suspect that I mess up the geometry during refinement. I have tried to
"tighten"
the geometry a bit
DELU $C_* $N_* $O_* $S_*
SIMU 0.05 $C_* $N_* $O_* $S_*
but this had little effect.
Has anyone refined against such data in shelxl? What refinement
protocol should be used and what are sensible values for the geometry restraints
at this resolution? I have thought about using NCS restraints, but
have not found a more detailed description (despite the shelxl manual and the
lysozyme refinement tutorial by Thomas Schneider) on how to define them,
given that my three chains are not really identical. If you'll put more then one
AFIX statement per chain, how does the program relate these groups? Finally,
it may be important to consider that each chain contains about 450 residues
making
refinement a run rather CPU/time consuming task: Is it still worth,
possible to try L.S. refinement?
thank you!
michael
p.s.: below I include the starting section of the last .ins file I have used.
TITL
CELL 0.93300 88.07 88.07 128.47 90.000 90.000 120.000
ZERR 3 0.088 0.088 0.128 0.00 0.00 0.00
REM Space group P 32
LATT -1
SYMM -Y, X-Y, 2/3+Z
SYMM -X+Y, -X, 1/3+Z
SFAC C H N O S
UNIT 13035 24267 3525 3663 48
DEFS 0.02 0.1 0.01 0.04
CGLS 10 -1
SHEL 12 2.47
FMAP 2
PLAN 200 2.3
LIST 6
WPDB 2
TWIN 0 1 0 1 0 0 0 0 -1
DELU $C_* $N_* $O_* $S_*
SIMU 0.05 $C_* $N_* $O_* $S_*
REM ISOR and CONN 0 recommended on adding water
BUMP
REM HOPE
REM Remove MERG 4 instruction if Friedel opposites should not be merged
MERG 4
Michael Hothorn
Structural & Computational Biology Programme
European Molecular Biology Laboratory (EMBL)
Meyerhofstrasse 1
69117 Heidelberg
Tel: 0049(0)6221 387 268 (office)
Tel: 0049(0)6221 387 609 (lab)
http://www.hothorn.de
