[ccp4bb]: map coefficients in Refmac and CNS

[Jorge Iulek <iulek@interponta.com.br>]

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

Dear all, 

  Well, I did take some time for RTFM, looking at sites, and archives but could 
not find exactly the point. 
Anyhow, references are welcome. 
  I am refining a structure with a ligand (AMP). The ligand density is, say, 
only reasonable (well, depending on the program...). 
Using CNS, the (u m F_c- v D Fo) phi_calc maps 
with u and v = 2 and 1 and u and v = 1 and 1 output by the program are pretty 
nice around the ligand. Ohterwise, when I use 

REFMAC, I calculate maps 
using FWT/PHWT (for the weighted 2Fo-fC style) and DELFWT/PHDELWT (for the Fo-
Fc style) (although refmac also outputs a PHIC), 

and these 
do not show a good density around the ligand, in fact showing breakages in some 
parts (as if the ligand were other 
roughly similar one ...). 
  So, questions: 
1) Refmac estimates m and D based on the R-free set and deals properly the 
absent reflections to avoid model bias on the map coefficientes

... Does CNS do this too ? I could not find this information in either the CNS 
documentation or 
in an old X-plor manual ... 
2) What is the so called phi_calc phase in CNS ? The corresponding PHIC of 
REFMAC, that is, based solely on the model coordinates  ?

 But, should these phases not bias the density 
towards the model 
 and then explain why I see a better density around my ligand in CNS maps ? 
3) Also, where could I find information on how PHIWT and DELPHIWT in refmac are 
calculated ? In the initial article (Acta Cryst.

D53, 240-255, 1997), it is annotated to calculate the maps with phi_c (phi in 
greek). When did those "new" weighted phases 

(PHWT and PHDELWT) were introduced (in refmac or wherever) ? 
        Resolution is  2.0. 
   Thanks for any response, point to literature, etc... 

Jorge