simtk.org

Hi,

I am using a 64 bit linux cuda system with the latest version of Gromacs.

Q1) Are all the force fields presented by pdb2gmx available to be incorporated on the gpu-gromacs?

0: GROMOS96 43a1 force field
1: GROMOS96 43a2 force field (improved alkane dihedrals)
2: GROMOS96 45a3 force field (Schuler JCC 2001 22 1205)
3: GROMOS96 53a5 force field (JCC 2004 vol 25 pag 1656)
4: GROMOS96 53a6 force field (JCC 2004 vol 25 pag 1656)
5: OPLS-AA/L all-atom force field (2001 aminoacid dihedrals)
6: [DEPRECATED] Gromacs force field (see manual)
7: [DEPRECATED] Gromacs force field with hydrogens for NMR
8: Encad all-atom force field, using scaled-down vacuum charges
9: Encad all-atom force field, using full solvent charges



Q2) I seem to be having trouble with the energy file, in that the file (*.edr) remains empty. Any reasons why that could be happening?

Thanks,
VJ

> Q1) Are all the force fields presented by pdb2gmx available to be incorporated on the gpu-gromacs?

No, some of them use functional forms that aren't supported by OpenMM. The only force field that is *fully* supported is AMBER, which you can get from

http://chemistry.csulb.edu/ffamber/

OPLS is *almost* fully supported. The only problem is that OpenMM uses Lorentz-Bertelot combining rules for Lennard-Jones interactions, not the Jorgensen combining rules that OPLS calls for. This will make a slight (but probably very small) difference in your results.

> Q2) I seem to be having trouble with the energy file, in that the file (*.edr) remains empty. Any reasons why that could be happening?

Make sure you're using PR4, since earlier releases didn't output energies. In PR4 the following energy terms should be reported correctly: potential energy, kinetic energy, total energy, and temperature. All other values in the edr file will be zero.

Peter

Thanks Peter,

We are working on the force field porting and hope to be completed with it shortly.

For the energy, I am using PR4, but may be I should double check my flag settings.

Thanks Again,
VJ