http://krystal.karlov.mff.cuni.cz/ecm/abstract/c/abstc4.htm
There is a link to 300 references (soon 302) in one of them.
Powder diffraction is a too narrow field for molecule
locators. I guess that single crystal data are excellent
for locating even much more giant molecules, if one
is not interested in refining without constraints. Imagine
having 30000 hkl and cells as large as 1000**3 Angstrom
cube. This will never be discussed in the powder
diffraction topic. This is why I suggested a distinct topic.
Those methods will not be specific to powder diffraction,
as well as direct or Patterson methods are not. But they
will be (and are yet) applied to powder data too, this is
evidence.
Armel
>K. Shankland, W.I.F. David, T. Csoka and L. McBride
>
>Structure solution of Ibuprofen from powder diffraction
>data by the application of a genetic algorithm combined
>with prior conformational analysis
>
>Int J Pharm 165 (1998) 117-126
>
>Essentially, the structure solution process is accelerated
>when torsional constraints derived from the Cambridge
>Crystallographic Database are employed. A related paper
>(Shankland et al, IJP 165 107-116) describes the prior
>conformational analysis in full detail.