[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[ccp4bb]: Re:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Try to use damping factor
DAMP 0.5 0.5
to reduce shifts at each cycle.
Garib
On Monday 09 Sep 2002 3:00 pm, rams@poori.biochem.uiowa.edu wrote:
> Hello All,
>
> I am refining a large sturcutre 1500AA/asym to 1.1 A resolution.
> I do isotropic looks good. Then TLS with 8 groups - R-factor and R-free
> drop over a percent.
>
> then I go ahead and do anisotropic refinement and things begin to go
> crazy.
>
> My R-factor and R-free at the end of TLS is 17.7 and 19.4.
>
> When I refine on my Linux box (ccp4 -4.2 ) this is what happens...
>
> $$
> Ncyc Rfact Rfree FOM rmsBOND rmsANGLE rmsCHIRAL $$
> $$
> 1 0.191 0.201 0.828 0.011 1.373 0.089
> 2 0.194 0.201 0.812 0.011 1.420 0.092
> 3 0.211 0.216 0.756 0.013 1.483 0.095
> 4 0.225 0.227 0.736 0.015 1.536 0.099
> 5 0.237 0.238 0.695 0.017 1.606 0.106
> 6 0.246 0.254 0.695 0.018 1.689 0.110
> $$
>
>
> With loads and loads of eigen value errors (too small -reset).
>
> However, on a note of desparation I tried to do the same on an XP1000 and
> I had the same problem - however when I swith to an SGI R10000, IRIX 6.5
> Something strange happens - everything looks great... - see below...
>
> $$
> 1 0.173 0.186 0.875 0.011 1.373 0.089
> 2 0.170 0.184 0.879 0.010 1.341 0.088
> 3 0.168 0.181 0.883 0.009 1.335 0.087
> 4 0.169 0.182 0.882 0.009 1.328 0.087
> 5 0.168 0.182 0.882 0.009 1.326 0.087
> 6 0.169 0.182 0.882 0.009 1.323 0.087
>
>
> Both jobs had exact same starting PDB file and exact same mtz file and
> exact same script file....
>
> so whats going on??????
>
> thanks.
>
> Rams.
--
Garib N. Murshudov, Chemistr Department, University of York
Tel: (01904) 328252