AMBER Archive (2002)

Subject: mol2_to_off

From: Matthew Randolph Lee, Ph.D. (mrlee_at_sd.lionbioscience.com)
Date: Fri Aug 16 2002 - 17:15:41 CDT

> > Has anyone experienced any problems with antechamber's
> > generation of prepc files (Cartesian prep input files)?
>
> Yes, prepc generation is broken; you must use prepi (or mol2 can be
> an attractive option as well, followed by "mol2_to_off").
>
> In principle, this should not be that hard to fix, but we are moving
> to abandoning prepc (or prepi) in favor of mol2, so it may never happen.

        The only problem with using mol2 files with mol2_to_off and
subsequently importing the lib file, currently, is that antechamber does
not seem to be capable of outputting mol2 files with the gaff atom types,
even with the explict '-at gaff' flag, which is where I view a great deal
of the value in antechamber lies, i.e. in the chemical perception.

        And, of course, the minor downside to using prepi files as opposed
to prepc files is that one has to figure out the vector to translate the
small molecule back to its original coordinates, after loading it into
leap. But, this is what I end up doing, because I am not familiar enough
with the rules used by antechamber/gaff to assign the atom types manually.

        It will be nice when either the '-fo prepc' flag gets fixed or
when antechamber will be changed to be capable of dumping gaff atom types.

Sincerely,
--Matt

*******************************************************************
Matthew Randolph Lee, Ph.D.

Computational Scientist phone: (858)410-6534
Lion Bioscience Inc. fax: (858)410-6665
9880 Campus Point Drive e-mail: matthew.lee_at_lionbioscience.com
San Diego, CA 92121
*******************************************************************