CNS map coefficient file format

Robert Campbell rlc at groucho.med.jhmi.edu
Fri Oct 6 11:02:28 EST 2000


Dear CNSers,

I've come across a question about the map coefficient file that is
(optionally) output by CNS when calculating an electron density map.

In order to use the online omit map calculation available in XtalView we
prefer to use the map coefficients and then FFT them within xfit.  As a
comparison, the maps have also been converted to brix format for display
in O and we tried Christopher Putnam's cns2fsfour (see his web page at:

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

The brix format maps from mapman and the fsfour map generated by
cns2fsfour look very much alike.  When I look at the coefficient file
from CNS I see two phase values.  An example of a few lines from a file
are:

 INDE    31   13    6 MAP_FOM=     0.296 MAP_PHASE=   267.636
                   MAP_SCALE=     0.419 MAP_COEFF=    16.506    87.636
 INDE    31   13    4 MAP_FOM=     0.266 MAP_PHASE=    33.779
                   MAP_SCALE=     0.419 MAP_COEFF=    13.581   213.779

There is a phase value after the "MAP_PHASE=" string and also in
the second value after the "MAP_COEFF=" string.  Note that these two
phase values are 180 degrees apart.  Later on in the same
file one can see:

 INDE    31   11    9 MAP_FOM=     0.716 MAP_PHASE=   274.751
                   MAP_SCALE=     0.419 MAP_COEFF=    34.228   274.751
 INDE    31   11    8 MAP_FOM=     0.641 MAP_PHASE=   313.974
                   MAP_SCALE=     0.419 MAP_COEFF=     2.274   313.974

Here you may notice that the two phase values are identical.  In the list
of 28845 reflections, 15713 have different phases in a sigmaa-weighted
Fo-Fc map, whereas only 5166 have different phase values in
sigmaa-weighted 2Fo-Fc map. 

Can anybody tell me where the two different values come from?  I can tell
you that if you use the second value, then the map looks virtually
identical to that created from the original CNS map and put through either
cns2fsfour or mapman for xfit or O respectively.  On the other hand, the
"MAP_PHASE" value gives a map with rather large amounts of density at
3 sigma where atoms already exist.

Any assistance would be appreciated.

Cheers,
Robert
-- 
Robert L. Campbell, Ph.D.                           <rlc at groucho.med.jhmi.edu>
Research Specialist/X-ray Lab Manager       http://biophysics.med.jhmi.edu/rlc
Howard Hughes Medical Institute,            phone: (410)614-6313 FAX: 955-0637
Dept. of Biophysics and Biophysical Chemistry
The Johns Hopkins University School of Medicine,  
Baltimore, MD 21205-2185


---






More information about the X-plor mailing list