AMBER Archive (2009)

Subject: Re: [AMBER] Improper dihedrals between residues?

From: Bill Ross (
Date: Tue Jun 09 2009 - 14:04:02 CDT

> If I try to do so in the IMPROPER section of a given residue's entry
> in the *.prep file, leap complains about the nickel ion not existing
> in the residue (unsurprisingly). ...
> So, my question is: is it possible to manually specify an improper
> dihedral which crosses residue boundaries like this, and if so, how
> might I go about doing so?

The IMPROPER section in files has only one purpose: these
'impropers' are remembered by leap and regurgitated on saveamberparm
for comparison with the true impropers that leap derives by matching
atom types in the appropriate topology to the impropers in parm.dat
and frcmod files.

A long time ago, impropers were defined in the prep file and the
programs prep, link, edit and parm built everything with them.
Once leap was introduced, the method of matching impropers changed
to match that for bonds, angles and dihedrals (the only diff is
that leap allows there to be no improper). I added the printing
of impropers to aid in debugging leap and for people tuning
the improper parameters.

If you want to have an old-style improper span a residue, you will
have to program it in leap. Probably it is easier to user rdparm to
check which impropers from parm.dat/frcmod files are getting applied
than to worry about the legacy handling in leap.


AMBER mailing list