Dear x-plor runners,
The B-factor restraint discussion is very quiet. Please share your
experience as soon as you have tested the Tronrud
restraints. my experience was not as good as I expected.
I refined the B-factors with the default xplor values for asigma and bsigma
(asigma 2.5/2.0 for sidechain and mainchain, bsigma 1.5/2.0)
and then with the bsigma input from the experimental values. I also used
tha gaussian distribution around the bfactors with 1.5 for the main chain
and 2.0 for the sidechains.
Xplor values give me lower Rfactors than the tnt restraints. The tnt
restraints give a lower (by 0.3%) working R, but higher Free R (by 0.4%),
so it also yields to more difference between the two values. These values
are from such runs, where xplor is allowed to set the weight to the bsigmas
automatically. The weight is set very low (0.01) and if I set it higher by 0.1
steps, the working R increases as a function of weight. The free R has a
minimum at weight 0.1, it is higher than the one I got with the
xplor overall bsigma restraint input.
So my experience wasn't as good others seem to have. My data is to 2.2 A
and collected at -15 C. I'm also interested to hear other peoples
experience for I also do think this was a very good idea, and am somewhat
uncertain why it doesn't seem to work with my data.
Pirkko Heikinheimo
--
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Pirkko Heikinheimo
Turku Centre for Biotechnology,
P.O.Box 123,
FIN-20521 Turku, Finland
e-mail: pirkko.heikinheimo at btk.utu.fi
fax: 358-21-333 8000
phone: 358-21-333 8024
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
