Re: AMBER: New (small) Topology without solvent from previous one (big) with solvent

From: Thomas Cheatham <>
Date: Thu, 13 Jul 2006 10:34:00 -0600 (Mountain Daylight Time)

> I would like to ask if there is anyway to get new topology file
> where solvent (or masked residues or atoms from ptraj stripped )
> from previous one which had included all these residues.
> The purpose of this questions is about ptraj use, where I'm getting
> average structures for masked residues stripping the rest.
> The output I want is a restart file and also a pdb.

It is not exactly clear what you want here, but the general answer at
present is "no". If I understand correctly, you would like to be able to
use the ptraj strip commands to strip arbitrary atom selections and then
write out a new prmtop that corresponds to the new (strip'ed) system.

In general, this is not a trivial problem since as atoms are removed, the
remaining atoms no longer represent the complete residue and the remaining
parameters may not meaningfully describe the (partial) residues.
Decisions have to be made about what the parameters should be. The
perhaps simplest implementation could be to assume that if an atom is
deleted, all internal parameters including this atom are removed (i.e. any
bond, angle, dihedral containing that atom is deleted). This could lead
to funny topologies (for example if the WAT.O atoms of TIP3P waters are
deleted, you will have a bunch of free hydrogens with no vdw floating
around) but would solve the problem of LEaP not recognizing the (modified)
residue from a strip'ed PDB file.

The only prmtop modification that is currently done in ptraj is changing
the number of solvent molecules and the resulting prmtop (which must be
written out) is currently written in old-style prmtop format. rdparm can
delete specific bonds, angles or dihedrals, however this also outputs the
old-style prmtop format and the commands are tedious. Given this, at
present you are stuck with LEaP or must make extensive modifications to

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Sun Jul 16 2006 - 06:07:02 PDT
Custom Search