Hello,
I would like to determine a structure containing a molecule whose chirality is uncertain at certain carbon sites: is it S or R? So I need to swap atomic groups (e.g. a proton and a phenyl group) rather than atoms during the structure solution. swap_sites is thus not of much help, since it just swaps two atomic sites.
The molecule is defined as a rigid body in a z-matrix description, so if properly defined, one should just swap two different torsion angles, or constraint one torsion angle to two different values - which is more reasonable for the example with a proton and a phenyl group.
The way I do it - and it seems to work - is as follows:
prm !a1 181.7
prm !a2 301.4
penalty = IF ( p1 < (a1+a2)/2.0 ) THEN
(p1 - a1)^2
ELSE
(p1 - a2)^2
ENDIF;
where p1 is the torsion angle which can take two different values a1 and a2
Does anyone have a better solution? How do you really swap two different torsion angles?
Arie