[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ccp4bb]: CCP4 Patterson Map -> XtalView Map



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


On Wed, 1 Mar 2000, Paul Hubbard wrote:

> Does anyone know how to convert a CCP4 patterson map to one which can be
> viewed in XtalView. Thanks.


Also check out my conversion utility ccp2fsfour -- and cns2fsfour -- (as
advertised on the XtalView web page):

http://www.scripps.edu/~cdputnam/software/ccp2fsfour.html

also

http://www.scripps.edu/~cdputnam/software/cns2fsfour.html 

for CNS/Xfit users which is finally a self-contained program rather than a
script which ran mapman-mapmask-ccp2fsfour).

I've also got a FSFOUR to CCP4 map file converter on the way (fsf2ccp4).
It currently writes out map files readable by CCP4 programs, but I haven't
had a chance to verify the output directly with a graphics program.

Note that while CCP4 map files have machine type stamps, FSFOUR files
don't.  That means that mixing alpha/intel maps with sgi/sun is a bad
idea.  It's best to convert ccp4 maps to fsfour on the machine that you 
are going to be running xfit or xcontur on.

The only current limitations of the software (which happily aren't an
unholy mixture of C and Fortran) is that they don't expand the asymmetric
units/molecule extents into the entire unit cell as demanded by the FSFOUR
format.  See the documentation, but easy solutions exist to solve these
problems (e.g. mapmask for CCP4 and map extents in CNS).  The program will
properly convert a map that doesn't cover a unit cell, but it flattens all
areas without any information to zero.  Additionally, the current software
reads in and then completely ignores the skew information in the map
header.  (Is there demand for supporting these options?  Do any CCP4
programs actually output a skewed map?)

For whatever reason, we typically find that the density in CNS maps
directly converted to FSFOUR are superior to those calculated from the
structure factors that CNS can output even when the finer CNS sampling is
taken into account. (Finally traced a difficult loop after the program was
written.)  We haven't come up with any reasonable explanations for these
observations.


On Wed, 1 Mar 2000, B. Hazes wrote:

> Just create a "normal" structure factor file for XtalView, but with the
> amplitudes set to your patterson coefficients (F-squared), the phases
> set to zero, and the FOM set to 1.0. Xfit will then calculate the
> Patterson for you.


That'll work.  I'd actually suggest just making a normal structure factor
file with F's and phases or a .fin file and pick the Fo*Fo (Patterson)
button.

Something that may not be clear from the documentation, but xfft will
happily read in .fin files (even though the input is labelled "phase
file") and calculate Pattersons with them. (If you issue the command xfft
mydata.fin mydata.map, xfft defaults to a Patterson).

Note that at the moment, .fin files must have 7 columns:

H K L  FP1 SIGFP1 FP2 SIGFP2

with missing reflections being indicated with FP=0.0 and SIGFP=9999.0 and
H K L can't have decimal points.


Phase files have 5 columns and are interpreted in one of two different
ways depending on the option chosen:

H K L Fobs Fcalc Phase

or 

H K L Fobs FOM Phase





Chris Putnam
------------------------------------------------------------------------------
                                                The Scripps Research Institute
                                                         MCSC Graduate Student
                                                          cdputnam@scripps.edu