*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
I have read the message 'how to convert CNS density modified file
to mtz forarp/warp' from ccp4bb, it is quite useful, and now I
have successfully convert .hkl to .mtz.(the method I used is SAD)
But when I do auto-build in the ARP, I got trouble!
in the ARP, there are 4 labels: 'Fobs, Sigma, PHIB, FOM'.
But in my .mtz file I cannt find 'Sigma', and when I run the
program, I was told that the label of 'Sigma' should be added. So
could you tell me what should I do and where I can find 'Sigma'?
(I can't find standard deviation at .hkl file coming from density
modification as well)
Thank you very much!!!
Best
Hao
In case you do not have an initial mtz file somewhere with the Sigma
(standard deviation) values --- to be merged with your current file
assuming that the relative scale factor has been maintained ---
then you
can easily compute 'fake' Sigma values as Sigma = 0.1 * Fobs.
Easily done using e.g. Bart Hazes's sftools.
The resulting values will suit most needs.
Fred.
On the other hand, there is no good excuse to miss the sigma's, so
better make
sure you recover from a file, together with the Fobs (so both are in
the same scale!)
and then use CAD (Merge MTZ files) to put together in the same file
Fobs/Sigma from wherever you have them, and the density modified Phi
and fom from CNS.
I would also calc a quick map with these Fobs and PHIs to make sure
that it looks reasonable
and nothing went wrong in format conversions.
Alternatively give up with CNS and use CCP4 or PHENIX; I am quite
puzzled why anyone would use
CNS these days that *** the CNS team themselves have released their
new software suite, PHENIX, which is free for academics ***, let
alone using CCP4 software (ehm ... plus SHARP) which is even better
than the ccp4bb itself ;-).
Using files from PHENIX.Resolve in ARP/wARP is completely
straightforward (they are mtz), let alone that phenix.resolve (and
Phenix.textal) model building might do the thing and you might not even
need to use ARP/wARP (with the emphasis on *might* and not on the
*not*, ARP/wARP is always good for you ...)
A.
