Re: [ccp4bb]: refining two conformations related by crystallographic symmetry

[Edward Berry <eaberry@lbl.gov>]

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

Summary of responses to my questions of last week -
(hope the answers weren't obvious as a pikestaff to everyone but me)

Question 1. What program can positionally refine overlapping sym-related atoms 
	at 0.5 occupancy?

Bernhard Rupp and Christine Cardin reported that SHELXL can do this.

Eleanor Dodson said REFMAC will not invoke VDW clashes 
	between symmetry equivs where both occs = 0.5.

Weiru Wang showed me how to use CNS and modify the selection of atoms 
for which interactions are checked. In the minimize script, well down 
below the line that says:
{===========================================================================}
{        things below this line do not normally need to be changed          }
{===========================================================================}
you find this section: 

1    constraints
2      interaction ( store1 and not(attr store4 > 0))
3                  ( store1 and not(attr store4 > 0))
4      evaluate ($alt=1)
5      while ( $alt <= $nalt ) loop alcs
6        interaction ( store1 and ( attr store4 = $alt or attr store4 = 0 ))
7                    ( store1 and ( attr store4 = $alt ))
8        evaluate ($alt=$alt+1)
9      end loop alcs
0    end

You can modify line 2 to exclude the residue which overlaps with sym-related:
     interaction ( store1 and not(attr store4 > 0) and not (segid="HHHH" and residue 241 ))

then you will check interactions of everything except this residue with 
everything including this residue, but will not check interactions between 
the residue and itself. If you have residues in multiple conformations 
(store4 > 0) which interact with the special residue, you might need 
to exclude it in lines 6 or 7.  In my case just changing line 2 seems
to work. (setting store4=1 for the residue doesn't work- it and its symmetry 
mate are seen as being in the same conformation 1 and clashes occur in 6 and 7)

=====================================================================================
Question 2. How can I keep the R-free set valid when expanding to lower symmetry?

Eleanor Dodson- let CAD expand the Free-R flags together with the data