Re: getting different dihedral energy with two identical backbones

From: Jim Phillips (
Date: Fri Sep 26 2014 - 17:18:24 CDT

Hi Joyce,

Looking at the code in Molecule.C I can see that pairInteractionSelf isn't
going to behave as you expect for dihedrals etc. that contain both
selected and non-selected atoms. This a bug and it is easily fixable.

A workaround is that, since the latest VMD beta (at least) can write full
psf files, you can load each structure into VMD, make a selection of the
backbone atoms, and write those selected atoms out as psf and pdb files
that you can then load into NAMD for energy evaluation.


On Wed, 24 Sep 2014, Joyce Yang wrote:

> Hi,
> I used the pairinteractionself function to calculated energy of the
> backbones of two structures (one is obtained from mutating a single residue
> with vmd plugin), respectively. The settings are as below,
> "pairInteraction on
> pairInteractionGroup1 1
> pairInteractionFile namd-temp.pdb
> pairInteractionSelf on
> coordinates namd-temp.pdb
> "
> The beta-coupling column of backbone atoms are set to 1 in the pdb files.
> Surprisingly, the calculated dihedral angle energy of the backbones are
> different in over 100 kcal/mol between the two structures. It doesn't
> really make any sense to me since the two backbones are identical (RMSD=0).
> Anyone has an explanation for this?
> Thanks
> Joyce yang
> Department of Chemistry
> University of Alabama at Birmingham
> Email:

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:52 CST