I’ll start a new chain on this point: Some JSME/JSmol users may not know the history on this point: Bob hooked Jmol to JSME in a useful way before anyone knew about Resolver! Specifically, he did some magic (circa 2008) with UFF and 3d information that is stored in the jme file. For small to moderate molecules, this load mode is still quite good.
I’ve been working to run as much of my app as possible off line - i.e. locally. To this end, I don’t even go to Resolver any longer for 2D/3D changes. The older pre Resolver approach is good enough for my small molecule app. For anyone who thinks they are getting “better" models from Resolver, check bond angles from time to time! Resolver was not created for model optimization. Far from it, it was created as a quick route to reasonable 3d files for cheminformatics purposes. Resolver is primarily a cheminformatics machine, and a beautiful machine it is. Its algorithm, however, does not really involve calculation. It simply detects sub-units called ensembles and puts them together from prefabricated molecular ensembles to form the molecule. This prefab approach is evident from consistency of aforementioned bond angles. If you want calculation optimized models, go to PubChem. They are MMFF94 optimized. Otis -- Otis Rothenberger o...@chemagic.com http://chemagic.com ------------------------------------------------------------------------------ _______________________________________________ Jmol-users mailing list Jmol-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jmol-users
