<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Morten,<br>
<br>
- when optimize_wxu=true is used the weight may be any that gives the
best Rfree. The optimized value is used only for a given run and not
propagated to .def file.<br>
- if "tls+individual_adp" is used (and optimize_wxu=false) then the
weight is set up automatically to a value found from systematic
re-refinement of models from PDB.<br>
<br>
Pavel.<br>
<br>
<br>
On 4/8/09 4:36 AM, Morten Grøftehauge wrote:
<blockquote
cite="mid:4018faab0904080436t2e5955a8l5ba751cd8880c0d8@mail.gmail.com"
type="cite">Just a quick question/comment,<br>
<br>
I may remember this incorrectly but isn't the best wxu_scale=1.81 when
you use TLS? Older versions of phenix.refine changed it automatically
when you applied TLS IIRC but now it only goes to 1.81 if I use the
optimize_wxu option.<br>
Am I remembering correctly or is my headache screwing with me?<br>
<br>
Sincerely,<br>
Morten<br>
<br>
<div class="gmail_quote">2009/1/26 Pavel Afonine <span dir="ltr"><<a
moz-do-not-send="true" href="mailto:pafonine@lbl.gov">pafonine@lbl.gov</a>></span><br>
<blockquote class="gmail_quote"
style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi
Scott,<br>
<div class="im"><br>
> In general should one optimize wxc and wxu ONLY as a final
polishing<br>
> step or can this be done earlier in the refinement and then the<br>
> optimized values used for subsequent rounds of refinement?<br>
<br>
</div>
Optimizing weights is time consuming: depending on the data and model<br>
size it can take from many minutes to hours, and this is the only reason<br>
for suggesting it as a final step.<br>
<br>
However if it runs quickly or/and you are not in rush, you can do it all<br>
the time: then it will always make sure that the weights are "optimal"<br>
(in terms of giving the best Rfree). Note, that most of the time you<br>
don't really need it since the automatically determined weights are<br>
pretty good (most of the time).<br>
<br>
The optimal weight found after running phenix.refine with<br>
optimize_wxc=true (or/and optimize_wxu=true) may substantially vary<br>
between structure determination stages or even between refinement<br>
macro-cycles within one refinement run. This is why it is not generally<br>
advised to use the same weights (determined at the beginning) all the<br>
time, although technically you can: just pick up the best wxc_scale from<br>
the log file and specify it for the next run.<br>
<br>
Please let me know if you have any questions!<br>
<font color="#888888"><br>
Pavel.<br>
</font>
<div>
<div class="h5"><br>
_______________________________________________<br>
phenixbb mailing list<br>
<a moz-do-not-send="true" href="mailto:phenixbb@phenix-online.org">phenixbb@phenix-online.org</a><br>
<a moz-do-not-send="true"
href="http://www.phenix-online.org/mailman/listinfo/phenixbb"
target="_blank">http://www.phenix-online.org/mailman/listinfo/phenixbb</a><br>
</div>
</div>
</blockquote>
</div>
<br>
<br clear="all">
<br>
-- <br>
Morten K Grøftehauge<br>
PhD student<br>
Department of Molecular Biology<br>
Gustav Wieds Vej 10 C<br>
8000 Aarhus C - Denmark<br>
Phone: +45 89 42 52 61<br>
Fax: +45 86 12 31 78<br>
<a moz-do-not-send="true" href="http://www.bioxray.dk">www.bioxray.dk</a><br>
<br>
<br>
<pre wrap="">
<hr size="4" width="90%">
_______________________________________________
phenixbb mailing list
<a class="moz-txt-link-abbreviated" href="mailto:phenixbb@phenix-online.org">phenixbb@phenix-online.org</a>
<a class="moz-txt-link-freetext" href="http://www.phenix-online.org/mailman/listinfo/phenixbb">http://www.phenix-online.org/mailman/listinfo/phenixbb</a>
</pre>
</blockquote>
</body>
</html>