[gmx-users] gromacs.org_gmx-users Digest, Vol 179, Issue 56

[jonathanjanke at gmail.com]

Hi Louise,

Aside from a forcefield compatibility problem, it’s useful to keep in mind that GLYCAM sometimes uses different 1-4 scaling factors for each dihedral, and AMBER is designed to run with this type of set up. GROMACS, on the hand, uses a “fudge factor” where all the 1-4 interactions are scaled by the same amount. I suspect this explains your ParmEd error. 

You (or someone else in this mailing list) may find a clever workaround for GROMACS to function with variable 1-4 scaling, but, unless you tweak GROMACS or your GROMACS-formatted parameters, the simplest approach would be to run GLYCAM within its native AMBER environment. 

If you check your AMBER-formatted parameters carefully and find that your particular GLYCAM parameters were designed with constant 1-4 scaling, you should be safe to run in GROMACS without any further modifications (I would still double check the GROMACS parameters by comparing single point energy calculations for both the AMBER and GROMACS parameters.). Otherwise, if you run GROMACS-formatted parameters without ensuring that any original 1-4 scaling factors are handled properly, running GLYCAM with constant 1-4 scaling in GROMACS may unpredictably affect your GLYCAM results.

Best regards,
Joel Janke

> On Mar 21, 2019, at 5:22 PM, gromacs.org_gmx-users-request at maillist.sys.kth.se wrote:
> 
> compatibility between ff14SB and GLYCAM_06j-1 force    field