<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>fixing atoms</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Hello all,<BR>
<BR>
I am refining a structure which contains several uranyl sites. I know the atoms should be there based on peaks in anomalous difference fourier maps, peak heights are greater than 10 sigma. Likewise we have solved the apo-structure before with the same heavy atoms. If I perform simulated annealing, the heteroatoms get pushed as far as 8 angstroms away. The Asp(s) and Glu residues that coordinate the uranyl ion then just move into the region of the map where the U has vacated. The resulting fofc difference map has a huge peak(>11 sigma) where the U should be. I noticed that if I inflate the starting b-factor for the U, the positions will stay put(relatively). What is the best thing to do here? Can I fix their positions, so that this does not happen? How do I do it? I found a post in the archives that says i can use: refine.sites.individual="not element U". Will this work for the SA? I know it works during refinement. Anyway, the data is to 3 angstrom. Here is the result from some SA runs:<BR>
<BR>
following simulated annealing:<BR>
no u, starting b-factor for all atoms 20:<BR>
Final R-work = 0.2502, R-free = 0.2899<BR>
with u starting b-factor 20:<BR>
Final R-work = 0.2771, R-free = 0.3094<BR>
with u inflated starting b-factor for U:<BR>
Final R-work = 0.2682, R-free = 0.2983<BR>
After further manual rebuilding and TLS refinement(with u):<BR>
Final R-work = 0.2478, R-free = 0.2849<BR>
<BR>
<BR>
Also, during further refinement(if I reposition the U(s)), some of the U(s) still get displaced and the b-factors get extremely high, some as high as 500. I think this is due to them being displaced though. Other datasets which I have refined with phenix don't exhibit this behavior. Could it have something to do with NCS restraints on the protein?<BR>
<BR>
any suggests?<BR>
<BR>
Thanks in advance-<BR>
Todd</FONT>
</P>
</BODY>
</HTML>