GROMACS

A while ago I noticed significant changes in my simulation results between running gromacs 4.5.1 with the included amber-gs-s force field and gromacs 4.0.7 with Eric Sorin's amber ports. I eventually figured out that the parameter files in the amberGS.ff directory are identical (up to whitespace and comments) to those of the amber99sb force field. For example:

$ diff amberGS.ff/ffbonded.itp amber99sb.ff/ffbonded.itp
366c366
< CK CB N* CT 4 180.00 4.18400 2 ;
---

CK CB N* CT 4 180.00 4.18400 2 ;

Is it possible that the parameter files got mixed up? If I understand correctly, amber-gs-s should be amber-94 without backbone dihedral potentials and without 1-4 scaling.

I just downloaded Gromacs 4.5.4, and it doesn't look like anything has changed in that version.

Original mailing list post