[gmx-users] Free Energy Topology between A and B.

[Justin Lemkul]

On 5/29/16 3:37 PM, Nash, Anthony wrote:
> Dear all,
>
> I¹m a total newbie when it comes to Thermodynamic Integration, and until
> now I¹ve been happy with umbrella sampling. However, I¹ve found myself in
> a situation where I believe TI would be the most appropriate technique.
>
> I would like to determine the energetic contribution that a mutant amino
> acid has to a protein-protein interaction. The structure is very complex,
> so pulling along a reaction pathway is out of the question, thus an
> alchemical path might work.
>
> I wish to integrate from state A to state B. The number of atoms in the
> wild type amino acid of state A are fewer that those in the mutant amino
> acid of state B. The example given in section 5.7.4 of the manual
> demonstrate integrating from propanol to pentane, but due to the GROMOS-96
> FF each topology shares the same number of atoms. Atom PC2 and PC3 of
> state A does not require redefining in typeB chargeB and massB, they are
> transferred across as they are in both structures. How does one define a
> mixed A B topology when the number of atoms differ? I.e., if those two
> atoms in the above example were not needed in state B.
>

You need to define transformations between dummy and real atoms.  Look into the 
pmx program (this was mentioned on the list just the other day for an identical 
problem).

-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

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