molecular similarity



 Dear CCLers,
 I am looking for a method, or a software packet, that compares a
 "drug-like" molecule against a database of other such molecules. The
 comparison should be based on 3D structure and take into account
 conformational flexibility of the molecules. Additionally, I would like
 to use this together with a virtual database, i.e. one that consists of
 fragments that are assembled according to certain rules. It seems that
 the combinatorics of all that make compute time be an important issue.
 I have found a number of references in the literature that offer partial
 solutions (SEAL, FLEXS, PLM, 3D-MCSS), so I would like to ask:
 - what is used by practioners in the field,
 - how they compare against each other (is there a relevant review?),
 - have I maybe missed the one method that does it all.
 Best regards...
 W. Huber
 hubi { *at * } fdm.uni-freiburg.de
 Freiburg Center for Data Analysis and Modelling
 Germany