Xavier Gomis Rueth wrote:
>> I am trying to refine a twinned DNA structure with two alternative strands of
> different sequence and occupancy 0.5. I 'd like to constrain "ncs-like" the
> backbones of both molecules, but it fails as the sequence is not
> the same nor the number/types of atoms. Any ideas to overcome this ?
>> Thanks a lot in advance,
>> Xavier
X-plor refuses to apply NCS unless the atom and residue names are the
same. (It probably should just issue a warning.) The only way to use
NCS correctly is to rename the atoms & residues to match. For example,
if you are just applying NCS to equivalent backbone atoms, you could
just call every residue NUL. You'll of course want to save the actual
names, perhaps in STORE1.
One other (and rather convoluted) possibility would be to generate an
extra copy of one chain with an occupancy of zero, and interactions
turned off, combined with strong NOE style distance restraints of zero
distance to the alternate chain. Then, just apply NCS between the real
and "invisible" versions of the first chain.
Good luck, Joe Krahn
