# Re: [AMBER] A clear procedure to do WHAM based on the REMD simulations outputs

From: Dongshan Wei <dswei0523.gmail.com>
Date: Sat, 31 Jul 2010 15:19:00 -0400

Hi Dan,

This week I first read the Chodera's code and tried to follow his
method, but eventually I gave up since I'm not familiar with python
and the installation of numpy and scipy also have problems in my
computer. Then I changed to use Grossfield's method, but after I read
the manual I found the code has the stability problem due to the
calculation of exp(-E/kT). Furthermore, both of these two methods
spend much effort on calculation of the standard deviation of
possibility (or PMF), which makes the program more complex. If we
don't care the standard error, I think the WHAM can be much simpler.
Now I want to write my own code. Before I begin, I have several

(1) How to determine the energy bin? Here the energy is total energy
or just potential energy in temperature REMD? In order to get the
energy distribution, whether or not are all of the energies from
different temperatures and different replicas counted?

(2) To read the energy values from T-REMD simulations, should I read
them from .mden file or .mdout file? Both of these two files records
the energy values at each interval of ntwr and ntpr. In addition, I
found the temperature for each replica at the same simulation time in
.mdcrd file and in .mdout is the same, but is different from any value
of Temp, T_solute, T_solv in .mden file. Why is there difference of
the recorded temperature in .mden and .mdout files?

(3) About the equation of g=1+2*tau, is it safe to simplify g=1? How
do you determine the correlation time tau and what is the magnitude in
order of tau for small biological system (one peptide chain and ~1000
TIP4P water)?

Thanks so much for time!

Best!

Dongshan

On Tue, Jul 27, 2010 at 7:34 AM, Daniel Sindhikara <sindhikara.gmail.com> wrote:
> Carlos,
>   The Chodera code is specifically made for T-WHAM. Alan Grossfield has
> implemented
> T-WHAM a year or two ago. HOWEVER, I had problems with its numerical
> stability for T-WHAM.
> I believe this is because it handles the DOS in normal space rather than log
> space (last I checked).
> --Dan
>
> On Tue, Jul 27, 2010 at 8:18 PM, Carlos Simmerling <
> carlos.simmerling.gmail.com> wrote:
>
>> do those codes do T-WHAM?
>>
>> On Tue, Jul 27, 2010 at 7:15 AM, Daniel Sindhikara <sindhikara.gmail.com
>> >wrote:
>>
>> > Dongshan,
>> >   You can either develop your own code, or use someone else's such as:
>> > Alan Grossfields: http://membrane.urmc.rochester.edu/content/wham
>> > or
>> > John Chodera's: http://www.dillgroup.ucsf.edu/~jchodera/code/WHAM/<http://www.dillgroup.ucsf.edu/%7Ejchodera/code/WHAM/>
>> <http://www.dillgroup.ucsf.edu/%7Ejchodera/code/WHAM/>
>> >
>> > Someone else might have a "more straightforward" way of doing this.
>> > I use my own code, but in its current condition I couldn't lend it out
>> > *yet*in
>> > full conscience. :P
>> > --Dan
>> >
>> >
>> > On Tue, Jul 27, 2010 at 7:43 AM, Dongshan Wei <dswei0523.gmail.com>
>> wrote:
>> >
>> > > Hi Everyone,
>> > >
>> > > Recently I want to do a ramachandran plot for an Alanine peptide. I
>> > > did a 10ns temperature REMD simulations with 16 replicas. Now I want
>> > > to weight the probability for temperatures not at T0 to get the
>> > > weighted ramachandran plot. I made a quick look through the mailing
>> > > list and find most WHAM is combined with umbrella sampling. Can
>> > > somebody give a clear procedure on how to do WHAM based on REMD
>> > > simulations outputs?
>> > >
>> > > Thanks so much!
>> > >
>> > > Dongshan
>> > >
>> > > _______________________________________________
>> > > AMBER mailing list
>> > > AMBER.ambermd.org
>> > > http://lists.ambermd.org/mailman/listinfo/amber
>> > >
>> >
>> >
>> >
>> > --
>> > Dr. Daniel J. Sindhikara
>> > Institute for Molecular Science
>> > E-mail: sindhikara.gmail.com
>> > --
>> > _______________________________________________
>> > AMBER mailing list
>> > AMBER.ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber
>> >
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
>
>
> --
> Dr. Daniel J. Sindhikara
> Institute for Molecular Science
> E-mail: sindhikara.gmail.com