This is actually two different questions. Both will be addressed here.

Question 1:  In a simulation, which protein force fields behave well with GLYCAM?

The GLYCAM force fields are designed to be independent of other biomolecular force fields, and, therefore, to work at least reasonably well with any of them.  The philosophy is that if each force field does a good job of simulating the behaviors of its own molecular components in standard, biological contexts, then the different force fields should be able to work together in those contexts.

That said, some form of validation should be performed on any simulation.  Since it is likely that any given researcher is using MD to study some previously un-studied molecule, then it isn’t possible to predict with absolute certainty how the simulation will perform.  So, always validate any combination of forcefields and simulation conditions.

It is also important to note the biomolecular context for which the force field was developed.  For example, the water model is a significant part of the environment. If it was also a significant part of the development of a force field, which is not necessarily true, then the force field might work well with only that water model.

Question 2:  Which GLYCAM force field files can I use with which of AMBER’s ff series?  Why?

This discussion does not consider the GLYCAM EP (“extra points”) force fields.

Details of changes between GLYCAM force field file versions can be found in the documentation on our parameters page.  But, here is a brief overview:

Avoid using GLYCAM file versions prior to 06h

Sometime circa 2011, it was discovered that the same atom type was being used in the GLYCAM files and in the AMBER ffx files.  This happening was accidental, and not desirable.  So, the AMBER developers settled on an atom-type naming convention designed to keep the various force fields from overlapping.  The 06h files were the first to employ the new naming convention.  Using older versions with a protein force field might result in protein parameters being applied to glycans or vice-versa.

GLYCAM files for AMBER ffs before and after ff12SB are not compatible

This only applies if a glycan is linking to an amino acid.  If the glycan is merely in the vicinity of the protein, and not covalently bound, the files are still compatible.

Starting with ff12SB, AMBER’s ffx force fields introduced new atom types needed to improve performance with amino acid side-chains.  GLYCAM provides amino acid files for covalent linkages between glycans and proteins (glycosylation).  Since glycosylation occurs on amino-acid side-chains, the GLYCAM files had to be updated as well.

Library (.lib) files for linking to pre- and post-ff12SB amino acids are available, but they must be used with the proper GLYCAM prep file (.prep) databases.  The only difference between the library files is in the assigned atom types.  Also, the more recent GLYCAM parameter (.dat) files retain parameters appropriate to the older ff’s, so it should only be necessary to use the proper lib files even if a more recent parameter set is used.

The following table summarizes. The most recent available files are recommended. All files can be obtained from our parameters page.