[gmx-users] PMF and RDF

Erik Marklund erik.marklund at chem.ox.ac.uk
Tue Dec 9 15:12:01 CET 2014


On 9 Dec 2014, at 11:00, Erik Marklund <erik.marklund at chem.ox.ac.uk> wrote:

> Dear Sujith,
> 
> Umbrella sampling does exactly that, but adds a biasing potential to sample high-energy regions of the reaction coordinate in separate simulations. The -px output when you do umbrella sampling with mdrun is indeed a sampling of distances along the reaction coordinate, which if you histogram it is the rdf. g_wham uses that data (or alternatively the force along the reaction coordinate instead of the distance) to build a PMF. My question to you is why do you think rdfs are significantly easier than pmts?

To clarify what I meant, in light of what Justin wrote earlier, is that the pmf and ref can *in principle* be calculated from exactly the same data. The *conventional* way to get the pmf involves biasing the simulations to sample poorly populated regions, whereas the rdf is commonly inferred from simulations without such bias. In principle the rdf can also be obtained from biased simulations if that bias is corrected for during analysis, and similarly a pmf can be obtained without bias. As Justin said, there are many situations where the sampling is insufficient for some distances however.

Erik

> 

> Kind regards,
> Erik
> 
> On 9 Dec 2014, at 06:39, sujithkakkat . <sujithks58 at gmail.com> wrote:
> 
>> Dear all,
>> 
>>    I read in *Phys. Chem. Chem. Phys., 2009, 11, 10427-10437*, that the
>> radial distribution function is directly related to Potential of mean force
>> through RDF=exp(-PMF/kT).
>> 
>>    My question is why would someone worry about computing PMF in a simple
>> case like interaction between two small solute molecules in water , along
>> the intermolecular distance, when one can get the RDF between the solutes ,
>> which I believe is easier than PMF calculation.
>> 
>>     Another article *Biophysical Chemistry 101-102 (2002), 295-307 *reports
>> PMF between solute molecules from Monte Carlo simulations. Why not just
>> find RDF.
>> 
>> Regards,
>> 
>> Sujith.
>> -- 
>> 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.
> 
> -- 
> 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.



More information about the gromacs.org_gmx-users mailing list