Re: [ccp4bb]: REVISE/ECALC//RANTAN/FFT

["Eleanor J. Dodson" <ccp4@ysbl.york.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

naismith@st-andrews.ac.uk wrote:
> 
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
> 
> Dear All,
>         We have stumbled upon a difficult MAD solution (or non solution as yet).
> We are experimenting with RANTAN. Can I check the the FM_RE calculated
> by REVISE is converted to EM_RE by ECALC which is on its own fed to
> RANTAN.
> 
> To find heavy atom sites, one then calculates a map using
> F1=EM_RE PHI=PHIbest_rantan and peaksearches it.
> 
> What I am unsure of is the EM_RE quantity, although derived from anomalous
> diffs, it is entered into FFT as F not DANO.
> 
> While I am on, does anyone have any thoughts or advice about what to play
> with RANTAN and what constitutes something to get encouraged by?
> 
> best
> Jim
> 


 Are you using the GUI - we hope that this guides you through the
various tasks..



  General points:

  REVISE generates an estimate of the true F_anom-scatterers ( called FM
or FA depending on your taste..)

 This is only as good as a) the measurements
                         b) the estimates of f' and f"

 I check all this quite neurotically using the scaleit summary generated
in the GUI.

 ( By the by, to do this you need to have set the project and dataset
names in the TRUNCATE step; to track the different wavelengths data sets
properly)

 Once REVISE has generated an FM, and ECALC the corresponding EM you can
use this as you see fit.
 
 It is sensible to look at the Pattersons this generates, but not
essential..
 It is rarely ( never?) necessary to use data past 3A to find the ha
sites, and certainly it can work at 3.5A or 4A. You should be governed
by the quality of the data..

Either run the rantan task which will read EM, generate phase sets and
then calculate maps and do peaksearches.

 Or run SHELX, taking FM as input ( SHELX dos its own normalisation)
either in the patterson search mode or the direct methods mode.

 Or run RSPS.
SLEX and RANTAN give ha lists ready for mlphare.

 Then do quick and dirty 3A? 4A? tests with mlphare centric refinement
only; the good solutions will give the lowest CullisR and useful
difference maps..

 Eleanor
> U.K., KY16 9ST                   | www     http://speedy.st-and.ac.uk/