[gmx-users] Free energy in octanol and water

[Gmx QA]

Hi Justin

Thanks, that is a good suggestion, I will try that.
Besides running multiple replicates, is there something else that is
obviously wrong with my method?

Best
/PK




2015-09-29 19:09 GMT+02:00 Justin Lemkul <jalemkul at vt.edu>:

>
>
> On 9/29/15 4:51 AM, Gmx QA wrote:
>
>> Hi gmx-users
>>
>> I am trying to calculate the free energy of solvation for a drug-molecule
>> in octanol and water.
>>
>> I am comparing two situations:
>>
>> A: Two systems, one with the drug in water only, and one with the drug in
>> pure octanol.
>>
>> B: drug in water only (same as in A), and then one box where I have added
>> 20% water to my octanol system (since mixtures of octanol and water
>> typically contains some water in the octanol phase).
>>
>> Then I run a series of free energy calculations per Justin's tutorial, 48
>> lambda points in total.
>> For system A I get a Delta-G for pure water of 39.07 kJ/mol, and 92.81
>> kJ/mol in pure octanol.
>>
>> For system B, the Delta-G for pure water is the same, but for the
>> octanol-box with 20% water the Delta-G is 95.99 kJ/mol.
>>
>> All of these results are obtained using gmx bar in gromacs 5.0.4.
>>
>> My main concern is that the Delta-G is higher in the water/octanol mixture
>> than in pure octanol, but I would have expected the reverse to be the
>> case?
>>
>>
> From experience, I can tell you it is very hard to get good convergence on
> such mixtures.  You'll note that in npt.gro you effectively have two bands
> of water formed in the box, neither or which is particularly close to the
> API molecule. You'll probably get some microscopic phase separation.  The
> route I've gone in the past is to generate several starting configurations
> (different random seeds when inserting the water molecules) and averaging
> over multiple runs.  You should also do some error analysis of your results
> (both block averaging for convergence, and to compare the two DG values
> between the systems).  Without error bars, you have no idea if those two
> values are at all meaningfully different.
>
> -Justin
>
> Is there something wrong with my protocol? The forcefield is
>> GAFF/amber99sb,and all production runs are 10 ns in length.
>>
>> Best
>> /PK
>>
>> I have uploaded copies of all input-files here:
>>
>> https://dl.dropboxusercontent.com/u/66785227/md.mdp
>> https://dl.dropboxusercontent.com/u/66785227/npt.gro
>> https://dl.dropboxusercontent.com/u/66785227/octanol_GMX.itp
>> https://dl.dropboxusercontent.com/u/66785227/API_gmx.itp
>> https://dl.dropboxusercontent.com/u/66785227/topol.top
>>
>>
> --
> ==================================================
>
> 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.
>