From: Mayne, Christopher G (cmayne2_at_illinois.edu)
Date: Fri Jul 26 2013 - 02:56:21 CDT

Daniel,

We're glad you are finding ffTK helpful!

1) Your understanding is correct, the dihedral potential can only be positive in the absolute sense. However, the PES obtained during the optimization is normalized to the global minimum for each particular PES, i.e., relative energies. This is required to make the QM and MM PESs directly comparable. Therefore, what is likely happening is that in the optimized PES, the global minimum has been raised, such that when the PES is normalized, the result appears to shift to lower energies. The important part being that the energy scale (y axis) is Relative Energy.

2) The latest version of ffTK (which have many important improvements and bug fixes) is available by downloading the latest alpha release of VMD 1.9.2. For instructions on how to obtain this development version of VMD, see:
http://www.ks.uiuc.edu/Research/vmd/alpha/

Regards,
Christopher Mayne

On Jul 26, 2013, at 1:29 AM, Daniel Klose wrote:

> Dear FFTK developers,
>
> thank you for the program, it's a great help!
>
> I've only got two brief questions:
> 1) In the resulting potential energy plot from the dihedral optimization, I observed that for some frames the final energy including the dihedrals has a LOWER energy than the initial MM energy without the dihedrals. The QM energy may be even lower.
> I thought from the equations that the dihedral energy can only be positive, so how can the final energy be lower?
>
> Since this is only the case for some frames, is that an indication that something went wrong for those few frames? The molecules looked normal though and there was no error in the QM calculation of the dihedral potential surface.
>
> 2) Is the newer paper on FFTK you mentioned in previous emails already available by any chance?
>
> Thanks for your advice on the dihedrals!
>
> With best wishes,
> Daniel