From: Wang Yi (
Date: Mon May 21 2012 - 08:47:10 CDT

> Hi
> On last February I asked (VMD mailing list) how to use fftk for
> transition metal enzymes. The developers kindly suggested what could
> be tried, pointing out, however, that fftk was not designed for
> TS-metal clusters. I did not succeed. I solved the task outside fftk,
> as I did previously for related cases, along two lines, one by simply
> using antechamber/gaff, with conversion to charmm ff through available
> software, the other one through R.E.D., getting a mol2/ff file and
> then manually arranging the frcmod file.
> Now, as charmm should better suited for a simulation I am aimed at, I
> am back to fftk for a small molecule comprising C,H,O,N,S atoms only.
> Therefore, it should be suitable for fftk Two questions:
> 1) In which order the various components of fftk should be run? I only
> have the PDB and a simple mol2 files, not comprising the FF.

Most likely you will find the sequence of the tabs is your sequence of working through.
When I started using it, I had a pdb and a psf. The topology and partial charge are produced based on analogy method from compounds that are already in the force field library. It's just an initial guess.
The first tab will tell you what FF parameters are missing (that you need to build), based on your psf file and the namd parameter file. So it's vital to correctly assign types to your atoms.

Then the general sequence is charge optimization, h-bond donor / acceptor optimization (the one I skipped), bond, angle, dihedral optimization.

> 2) At which level should QM calculations be carried out in view of the
> presence of sulfur atoms?

Sorry don't know.

> Thanks for advice
> francesco pietra