[gmx-users] Reproducibility in free energy calculations.

[Michael Shirts]

Mean of the three calculations is 4.35, standard error is 0.19.  So this
seems statistically consistent.  If you want to get more consistent
results, you'll need to run longer until the estimated statistical
uncertainty is lower.  Also, the estimated uncertainty is often an
underestimate (BAR often underestimates by 20-30%).


On Sat, Apr 19, 2014 at 9:05 AM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 4/19/14, 1:16 AM, sujithkakkat . wrote:
>
>> Hello Justin,
>>
>>     I tried the free energy calculations in your tutorial at different
>> sets
>> of conditions. Also I repeated a simulation at the same conditions with
>> same parameters, thrice on the same computer on same number of processors.
>> The results in the three cases where different (4.16 +/- 0.19 , 4.36
>> +/-0.14 , 4.54 +/-0.15 kJ/mol ) . I am not surprised, since I thought that
>> this might happen to a small system (241 water + 1 methane). However, I
>> want to be sure that this is OK. I guess in cases like this I should go
>> for
>> an average value of many runs.
>>
>>
> Offhand, I would have expected slightly better agreement between the runs,
> but there are tons of factors at play, so the outcome seems very plausible
> and reasonably consistent.
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 601
> 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.
>