[gmx-users] free energy settings

Justin Lemkul jalemkul at vt.edu
Mon Dec 5 00:10:47 CET 2016



On 12/4/16 3:09 PM, Qasim Pars wrote:
> Dear users,
>
> I have a few questions about the below free energy settings:
>
> ;State A
> free-energy              = yes
> init-lambda              = 0
> delta-lambda             = 0
> sc-alpha                 = 0.5
> sc-power                 = 1
> sc-sigma                 = 0.3
> couple-intramol          = no
> couple-lambda0           = vdw-q
> couple-lambda1           = none
> couple-moltype           = ligand
>
> According to the above settings the ligand in the state A will stay in the
> protein binding pocket. For state A the ligand has bonded and non-bonded
> interactions with the protein, right? Whereas I set/turn on only the
> non-bonded interactions (couple-lambda0 = vdw-q) but I couldn't find any
> command to turn on the bonded interactions for the state A. Is there any
> command to turn on the bonded interactions in GROMACS2016?
>

Well, there's bonded-lambdas, but this can't be the case here.  If your ligand 
is covalently bonded to the protein, then it isn't its own [moleculetype] and 
therefore can't be transformed in this way.

If you're computing the absolute binding free energy of a ligand to a protein, 
you don't want to change the bonded terms at all, otherwise your ligand will 
progressively disintegrate and you'll get a totally nonsensical result.

> ;State B
> free-energy              = yes
> init-lambda              = 1
> delta-lambda             = 0
> sc-alpha                 = 0.5
> sc-power                 = 1
> sc-sigma                 = 0.3
> couple-intramol          = no
> couple-lambda0           = vdw-q
> couple-lambda1           = none
> couple-moltype           = ligand
>
> As for state B the ligand doesn't have both bonded and non-bonded
> interactions with the protein, right? Therefore the ligand in the state B
> probably will move out from the protein binding pocket. Is this an expected
> result?
>

Yes, which is why you need some form of orientational restraints between the 
ligand and protein, then correct the free energy for the effect of these 
restraints.  This can be done with the pull code and 
[intermolecular_interactions] between the ligand and protein.

> I think for both two states the topology file can just be a normal topology
> as in a protein-ligand simulation. The ligand does not need to be a residue
> of the protein, and there is not need to have a state B defined in the

The ligand *can't* be a residue in the protein.

> topology file, as the couple- flags in the mdp already build the B state
> for the decoupled ligand automatically. Am I wrong?
>

This is correct.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================


More information about the gromacs.org_gmx-users mailing list