[gmx-users] pull code and binding free energy

[Justin Lemkul]

On 11/16/16 6:33 AM, Qasim Pars wrote:
> Dear users,
>
> I was trying to do the binding free energy calculation between protein and
> ligand. I get confused on the preparation of the input files:
>
> First question: Why we need to use the pull code in the mdp files for the
> binding free energy calculation (e.g.the pull code has used in this
> tutorial:
> http://www.alchemistry.org/wiki/GROMACS_4.6_example:_n-phenylglycinonitrile_binding_to_T4_lysozyme)?
> What does the pull code really do? I read the manual section 6.4 but it
> didn't answer the questions in my head.
>

When decoupling a ligand from a binding site, as the interactions get weaker, 
the ligand can sample unphysical states that inhibit convergence.  Various 
approaches, including distance and orientation restraints, are typically applied 
to preserve the binding pose and obtain better converged results.  Their 
contributions to the binding free energy are then straightforward to decompose 
to obtain the proper binding free energy.

> Second question: Instead of using the pull code in the mdp files, is it
> better to use [ bonds ] under [ intermolecular-interactions ] in the
> topology file?
>

No.  You need to use the pull code because its restraint potential is modulated 
as a function of lambda, allowing for the correction to the binding free energy 
to be computed.

> Third question: The PMF or the binding free energy (alchemical) is the best
> one to calculate free energy? And why?

Either should work.

-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

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