Re: [AMBER-Developers] [AMBER] Parmed: Residues cannot be part of 2 molecules! Molecule section will not be correctly set.

From: B. Lachele Foley <>
Date: Fri, 7 Aug 2020 06:34:39 +0000

I don't know how much the molecules are used in the MD calculations. But, they are used in the top files to differentiate solute from solvent. (info below for convenience) Lots of programs read prmtop files.

Do the IS method use Cartesian/bonded nearest neighbors to determine the effective radii or do they use molecules? I've always wondered that for other reasons.

Are molecules used in the coordinate wrapping for periodic conditions? I know the effect is to keep molecules together, but maybe the calculation follows bonds.

Even if no calculations/procedures use the info, I worry how many programs rely, in one way or another, on the information being in that format, at least while reading in the prmtop. I don't recall specifics, but I know I tried merely reordering the prmtop sections once and broke a program or two. That is - it might end up being quite a lot of work to implement the change throughout AMBER. But, I could be wrong. 🙂

Relevant prmtop section:

  IPTRES : final residue that is considered part of the solute,
           reset in sander and gibbs
  NSPM : total number of molecules
  NSPSOL : the first solvent "molecule"

Possibly, "molecule" could simply be turned into "residue". But, like I said, that could break other programs.

:-) Lachele

Dr. B. Lachele Foley (she/her/hers)
Associate Research Scientist
Complex Carbohydrate Research Center
The University of Georgia
Athens, GA USA
From: David A Case <>
Sent: Thursday, August 6, 2020 4:30 PM
To: AMBER Mailing List <>
Cc: <>
Subject: Re: [AMBER-Developers] [AMBER] Parmed: Residues cannot be part of 2 molecules! Molecule section will not be correctly set.


On Thu, Aug 06, 2020, Samuel A. Ratliff wrote:
>I've found an issue in my .prmtop file that is caused by creating a bond
>in tleap between a protein and its heme.

Modify the pdb file so that the heme group directly follows the protein chain
to which it will be attached.

[Amber developers: does anyone know which parts of the code still need to know
where the "molecules" are? I suspect we only have a molecular virial for
backwards compatibility with codes written 30 years ago, but I could be wrong
here -- I'm not really sure how the MC barostat works. Basically, I think
we have all these checks and atom-reorderings to deal with a problem that
really doesn't exist....]


AMBER-Developers mailing list
AMBER-Developers mailing list
Received on Fri Aug 07 2020 - 00:00:02 PDT
Custom Search