Hi,
You could turn down the weight on the B-factor restraints (in geometric parameters). It seems
like the restraints are two tight and do not allow the B-factor to increase enough to fit the
density.
~L~
_______________________________________
Lari Lehtiö
Structural Genomics Consortium
Medical Biochemistry & Biophysics Dept.
Karolinska Institute
Stockholm, Sweden
_______________________________________
----- Original Message -----
From: lithomas <[log in to unmask]>
Date: Monday, March 19, 2007 7:47 am
Subject: [ccp4bb] questions related to refmac refinement and map-fitting
To: [log in to unmask]
> Hello,
> I am not sure if I can clearly explain the problem , but ..
>
> The protein structure that I tried to solve is 407 a.a. in length.
> The crystal diffracts to 1.9 Ang and there is one molecule per asym
> unit. The program that I used is Molrep. Currently I have put in
> most of the residues and the R is 22.8% and the Rfree is 28.7% (no
> water added yet) after refmac (v. 5.2) refinement.
>
> I had a hard time trying to finish the refinemnt task because I
> noticed about ten residues' sidechains persistently had negative
> density. Typically what I would do is to put these a.a. back to ala
> and re-check the delFWT map in the next refinement cycle. However,
> the refmac delFWT map in the next refinement cycle would show
> strong positive (at least 2.5sigma) values. These amino acids don't
> seem to have alternate conformations and the Rfactor increase <1%
> and Rfree increase by 1%. Now, I seems to fall into this trap of
> endless loop and trying to find a way out.
>
> Since this is the first time that I used refmac, I am not sure if I
> did something wrong. Everything seems quite straightfoward from
> the pull-down menu and a couple main parameters that I selected are
> "restraint refinement", "no prior phase information", "anisotropic"
> and "babinet scaling". I also tried CNS (v. 1.1) and the results
> are pretty much the same.
>
> I know I should not let these residues bearing a large blob of
> negative density on the very end of the side chains. But did anyone
> have similar probelm and how to resolve this?
>
> Thanks for the help.
>
> Thomas
>
|