[gmx-users] REMD Statistics

[Mark Abraham]

On Mon, May 6, 2013 at 3:48 AM, Kong xq <xqkong776 at gmail.com> wrote:

> Hi Mark,
>
>
> Thanks for your great help.  I am sorry for the negligence to state the
> variation value correctly( it should be 0.011 rather than 0.11). Does this
> somewhat small value indicate the generalized equilibrium achieved?


It might be consistent with it, but it is never diagnostic of it. FWIW, I
have never seen an REMD study that attempted to address whether generalized
equilibrium was achieved.

I will
> search the papers you suggested. I am wondering whether the histogram is
> the gold standard for effectively constructing REMD flow. Moreover, how to
> do the potential energy overlap analysis in Gromacs, from the edr file for
> each replica ？Does any tool avalible in Gromacs to do this job ?
>

g_energy to get the data for each replica, then your favorite analysis
package to form histograms, and graphing program to overlay the plots. The
histograms can be done with g_analyze.

Mark