[gmx-users] Questions about free energy calculation tutorial

[gozde ergin]

Hi all,

I tried to estimate the free energy of betaine on betaine covered surface. My target molecule is also on the surface.
In order to do that I turn off charges linearly, then turn off LJ with a sort-core potential. 
Lambda is increased with 0.1 intervals for coulombic part and 0.05 intervals for LJ.
I did NVT simulation and I applied a flat-bottom potential on surface molecules.
However the result seems a little bit strange, first I could not see any minima in free energy from lambda 0 to 1 (I assume there should be minima aroma lambda=0.6-0.7). 
Also my histogram plot also seems strange that I have very big numbers on x axis (like e+33).
Please see the plots link below.

https://imgur.com/a/fqgmi

Any help would be appreciated.
Thanks.
> On 06 Nov 2016, at 14:33, Justin Lemkul <jalemkul at vt.edu> wrote:
> 
> 
> 
> On 11/6/16 8:27 AM, gozde ergin wrote:
>> 
>>> If it's the salt, then take the known Na+ value for the force field and add it to the SDS value you obtain at infinite dilution. Straightforward.
>> 
>> Ok either I can take hydration free energy of Na from the literature or I can run 2 thermodynamic integration with dodecyl sulfate and Na separately and sum them.
>> 
> 
> Yes, because DeltaG is a state function.
> 
>> Besides this I would like ask how did you manage the get the very similar solvation free energy result with experiments without adding Coulomb interaction which means decoupling only the L-J?
> 
> The reference quantity is just the LJ component of the total free energy; please see the referenced paper by Shirts et al.  Methane is net-neutral and the LJ will dominate, anyway, but the quantity shown (and explicitly stated) is NOT the hydration free energy, just part of it.
> 
>> Do you think to estimate the correct solvation free energy for dodecyl sufate(-) and Na(+) I need to decouple also the Coulombic interaction?
> 
> Of course, those will be critical for charged species.
> 
>> If yes, in order to do that I just need to modify the coul_lambdas parameter in .mdp file, right?
> 
> Yes.
> 
>> Also I should not make the charge of the atoms as zero in topology file.
> 
> Correct.
> 
>> According to the Sander Pronk tutorial of Solvation free energy of ethanol, I can turn off both the electrostatic (Coulomb) interactions and the Van der Waals (Lennard-Jones) interactions at the same time.
>> 
> 
> This may be unstable, and is generally not done.  Turn off charges linearly, then turn off LJ with a soft-core potential (or vice versa).  The point is you should never have charges on atoms with no (or drastically reduced) LJ terms. This can be easily specified in an appropriate series of lambda states in the .mdp file.
> 
> -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
> 
> ==================================================
> -- 
> Gromacs Users mailing list
> 
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
> 
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> 
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.