From: Md Shaik
Date: 14 September 2011 15:34
----------
From: Pavel Afonine
Hi,
----------
From: Phil Evans
Are you sure of the space group? P622 is much rarer than P6(x)22 where x > 0
Phil
----------
From: Tim Gruene
Dear Md. Munan Shaik,
a couple of aspects you might check:
- - first build the model AS MUCH AS POSSIBLE before running the first
refinement cycle
- - switch off automatic weight determination in refmac and use a low
weight (e.g. 0.005) and many cycles of refinement
- - check that the geometry does not distort (rmsd angles and bond
lengths reported in the refmac log-file must not go too high). if it
does, lower the weight even further.
if your model stems from higher resolution data and is actually in good
agreement with the model you are going at, refinement might ruin the
good model if you use inappropriate settings, causing Rfree to go up.
Cheers, Tim
Dr Tim Gruene
Institut fuer anorganische Chemie
Tammannstr. 4
D-37077 Goettingen
Date: 14 September 2011 15:34
Dear ccp4,
I am refining one structure at low resolution (3.5 A). I solved the structure in space group P622 by molecular replacement. The Rw/Rf after some cycle of rigid body refinement in Refmac5 is 34/37. But when I am trying to refine with Restrain refinement in Refmac the Rw is going down but the Rf is going up. This same happens with the phenix refinement. The protein is quite big 520 residues and one molecule is present in the crystal asymmetric unit. The maps are also not so bad except some region.
Please give me suggestion, what should I have to care during low resolution refinement? or is there any other tricks for this types of data?
Thanks in advance.
Md. Munan Shaik
PhD Student
Department of Biotehnology
School of Bioscience and Biotechnology
via G. Colombo 03
Padova 35131, Italy
PhD Student
Department of Biotehnology
School of Bioscience and Biotechnology
via G. Colombo 03
Padova 35131, Italy
----------
From: Pavel Afonine
Hi,
did you use refinement strategy suitable for low (3.5A) resolution (secondary structure restraints, Ramachandran plot restraints, proper ADP parameterization, SA, etc.... etc...)?
If you send me the data and model files off-list I will have a look.
Pavel
----------
From: Phil Evans
Are you sure of the space group? P622 is much rarer than P6(x)22 where x > 0
Phil
----------
From: Tim Gruene
Dear Md. Munan Shaik,
a couple of aspects you might check:
- - first build the model AS MUCH AS POSSIBLE before running the first
refinement cycle
- - switch off automatic weight determination in refmac and use a low
weight (e.g. 0.005) and many cycles of refinement
- - check that the geometry does not distort (rmsd angles and bond
lengths reported in the refmac log-file must not go too high). if it
does, lower the weight even further.
if your model stems from higher resolution data and is actually in good
agreement with the model you are going at, refinement might ruin the
good model if you use inappropriate settings, causing Rfree to go up.
Cheers, Tim
- --
Dr Tim Gruene
Institut fuer anorganische Chemie
Tammannstr. 4
D-37077 Goettingen
No comments:
Post a Comment