I would like to take a set of homologous molecules, "average" their
structures in some fashion to obtain a "model" that captures the most
conserved features of the structures. I intend to use such a synthesis
as a search model for molecular replacement of a protein with sequence
homology but unclear evolutionary relationship to any of the known
structures.
On short notice, I can think of some quick and dirty (and kind of
stupid) ways to try and implement this, but I am wondering if some
elegant formulation of this idea exists (ok, just one known to work!).
In particular, if there is a way to implement it under CNS. Thanks for
any help.
Charles E. Stebbins
