Lachlan efforts as CCP14 secretary in order to obtain (for us)
softwares for locating molecules in crystal structures were
fruitless, once again.
You have done your job, Lachlan, anyway, thanks. Some possibly
related programs are listed below. Try to obtain those which are not
commercial. No guarantee that this list is exact nor complete.
CERIUS for Model Building - commercial
DLS for Model Building (Baerlocher)
FOCUS (Mccusker. Grossekunstleve. Baerlocher, Yoshikawa, Davis)
GAP Genetic Algorithm (Shankland, David & Csoka)
subject of a patent application
GAPSS Genetic Algorithm techniques for Powder Structure
Solution (Johnston, Kariuki, Harris)
INSIGHT Biosym - commercial
MNDO (Bredas, Themans, Fripiat, Andre, Chance)
MODEL+PARAM (Byrom & Lucas) Determines molecular arrangements
which satisfy packing constraints
OCTOPUS96 Monte Carlo, used for fluorescein (Tremayne, Kariuki, Harris)
PATSEE (Egert, Sheldrick) rotation and translation search
to locate a fragment of known geometry
http://indy3.org.chemie.uni-frankfurt.de/egert/patsee.html
PMC (Dzyabchenko, Belsky & Zorkii) Packing calculation from atom-atom
potentials (AAP)
P-RISCON (Masciocchi et al.)
PROMET(5) (Gavezzotti) Generation of possible crystal structure
from the molecular structure
ROTSEARCH (Rius, Miravitlles) Rotating known fragments in a cell and
checking against intensities
ROTS96 (Rius, Ochando) Patterson search method
THEO Structure model generator, no ref. given
I am not sure I would be interested in solving structures fully
automatically (knowing yet the molecule), and possibly not being
able then to refine them. In the past I have abandoned EXAFS
because I was able to predict what interatomic distances
would be obtained with a precision of 0.01 Angstroms, before
to perform the experiment. Not very exciting !
So, I suggest to create a new topic, distinct from structure
determination, that would be named "molecule location".
This would be a world without surprise, where
benzene cycles could never be false, where hypotheses
would always be correct. The Rietveld method would not
really have a place in that world. Programs especially
conceived for rigid-body refinement would be clearly more
adapted.
In fact I am happy with inorganic materials and I don't
want to become a molecule locator. Even if I generally
know ineratomic distances, the tridimensional structure
discovery is a surprise which is my drug (as well as
chocolate :-).
Nevertheless, molecules should be located :). No matter
for a drama.
Armel