[gmx-users] Relative FEP topology question - What to do with state B dummy atoms in the .itp file under [ bonds ], [ angles ] and [ dihedrals ]

[Billy Noonan]

Hi Gromacs Users,

Just some background - I am doing FEP to determine the difference in
binding free energy between two ligands (state A and state B) wherein state
A has an isopropyl ether group and state B has a methyl ether group in its
place.  In other words I am transmuting the isopropyl group into a methyl
group.

I am using a united atom topology file for my ligands and the GROMOS 96
forcefield.  Ligand topologies come from the ATB.

The difference in dG between these two states should be +7.04, from
literature.  I am not currently getting this value, not even within an RMSD
of 10 kJ/mol.

In transfiguring the isopropyl (state A) into the methyl moiety (state B),
I have set the two CH3 and one H groups from state A into DUM atom types in
state B, and likewise transfigured the central carbon of the isopropyl
group, a C atom type, into a CH3 atom type.  I have re-adjusted the charges
on every state B atom for the state B ligand where necessary.

My question is this - how do I state that the [ dihedral ] angles, [ angle
] and [ bonds ] from the CH3, CH3 and H atom types annihilated, no longer
exist in State B, under the [ dihedral ], [ angle ] and [ bonds ]
sub-headings?

I have consulted the gromacs manual (section 5.7.4), but the table (table
5.6) and information included do not answer my question.  Or the table
(table 5.6) might have the information I need, but it is unclear what the
table is saying. The example provided is also too simplistic (propanol to
pentane) and I need to know what to do in my case.

I've been told that even non-interacting atoms can contribute energy by
virtue of their rotation so any advice on what to do would be greatly
appreciated.

Please let me know if you need more information.

Kind regards,



Billy