<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
A few more points:<br>
<br>
- define secondary structure and use it in refinement. If you don't
use secondary structure restraints then even ideal secondary
structure in starting model will deteriorate after refinement at
such low resolution. Most importantly, make sure that secondary
structure restraints that you defined are actually used in
refinement. If you suspect they are not used in refinement - please
report a bug right away.<br>
You can use phenix.secondary_structure_restraints to let Phenix
automatically identify secondary structure. It is critical that you
manually review it, and edit if necessary (which most likely will be
the case) in order to get the most accurate annotation!<br>
<br>
- while torsion NCS is a better option most of the time, at very low
resolution Cartesian NCS might be a betetr option. So, as Nat
pointed out, try both and see which one works best.<br>
<br>
- if real-space refinement does not perform well in your case,
please report a bug, otherwise use it as it will eliminate rotamer
outliers. Again, if there is any suspicion that it does not perform
as expected - please report a bug.<br>
<br>
- if there are Ramachandran plot outliers: fix them all manually
fist, then use Ramachandran plot restraints to prevent them from
re-occurring (Important: do not use Ramachandran plot restraints to
fix the outliers!).<br>
<br>
- if data set is severely incomplete - look at two 2mFo-DFc map:
regular and missing Fobs "filled" (phenix.refine outputs both maps).<br>
<br>
Good luck!<br>
Pavel<br>
<br>
<br>
<div class="moz-cite-prefix">On 4/26/14, 7:59 AM, Nathaniel Echols
wrote:<br>
</div>
<blockquote
cite="mid:CALeAa1Ngm0nwR1deWOGGKsteE-Hn8d6a6MJQyxGYynsMqwcWsQ@mail.gmail.com"
type="cite">
<div dir="ltr">On Fri, Apr 25, 2014 at 3:15 PM, Appu kumar <span
dir="ltr"><<a moz-do-not-send="true"
href="mailto:appu.kumar9@gmail.com" target="_blank">appu.kumar9@gmail.com</a>></span>
wrote:<br>
<div class="gmail_extra">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir="ltr">
<div> I want to refine the low
resolution structure (6.5A). It is a membrane protein
and we are able to phase it. It will be great help if
any one can guide me how to proceed further with the
refinement. People uses Phenix to refine low
resolution structure, so phenix is really a good
program to proceed for refinement against low
resolution.</div>
</div>
</blockquote>
</div>
<br>
</div>
<div class="gmail_extra">How to proceed depends on what exactly
you want the refinement to do. If the structure is already
approximately correct with no major conformational changes,
here are some basic guidelines:</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">1) Use tight restraints. The weight
optimization doesn't work very well in this resolution range
(at least in my hands), so setting very low values for
wxc_scale and wxu_scale (which control the weighting of the
X-ray term relative to the geometry and B-factor restraints,
respectively) is a good place to start. I would try values of
wxc_scale=0.025 and wxu_scale=0.05, but you may need to
experiment with this.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">2) Use NCS restraints if you have
multiple copies in the ASU. (The torsion restraints are
usually sufficient - and much easier to use - but it might be
worth trying cartesian restraints too.)</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">3) Use a reference model if you have a
good high-resolution structure available (this can be just
part of the structure).</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">4) Real-space refinement is probably
not going to work.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">5) Try using a very conservative
refinement strategy at first - for instance, rigid-body and
group B-factor refinement. The latter should work better now
(in the latest nightly builds, and imminent 1.9 release) since
Pavel added restraints on neighboring groups. You can also
try refining TLS for domains. My own experience was that
individual coordinate and B-factor refinement worked best -
the latter is somewhat controversial but with tight restraints
it ends up not being very different from refining an overall
B-factor. I also found it helpful to reset all B-factors to
20 at the start of refinement to remove bias from the starting
model.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">6) Be very paranoid about the 2mFo-DFc
map - at this resolution it will look almost exactly like your
model. Make lots of omit maps, and if you have any heavy
scatterers (metals, etc.) those will be very helpful in
confirming that you have the correct structure.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">If you actually have to do a lot of
rebuilding, it is considerably more difficult; DEN or Rosetta
refinement may be helpful.</div>
<div class="gmail_extra"><br>
</div>
<div class="gmail_extra">-Nat<br>
</div>
</div>
</blockquote>
<br>
</body>
</html>