[gmx-users] Can g_wham support using different temperature for different windows?

[Justin A. Lemkul]

Jianguo Li wrote:
> Thanks Justin.
> I tried your suggestions by either increase more windows and change the 
> force constant, but it seems the samplings are still bad in some 
> windows. When I did pulling in (0 0 1) direction and a reverse pulling 
> in (0 0 -1) direction, I got different configurations at certain 
> reaction coordinates. And the windowed umbrella sampling seems depends 
> strongly on the initial configurations in that window. Therefore I got 
> different PMFs using pulling in (0 0 1) direction and reverse pulling in 
> （0 0 -1) direction.
> 

How long are each of the simulations in each window?  Sufficient sampling should 
eliminate any configurational bias and/or hysteresis.  Also, if the pulling that 
sets up the initial configurations is done slowly enough, you won't see these 
problems.  Sounds to me like you're pulling too fast or hard, such that the 
system is not stable.

> In my simulation, I exert constraints on phosphate atoms in z direction, 
> so there is no lipid flip-flop and the membrane will be stable at high 
> temperatures. Then I am thinking of increasing temperature in those bad 
> windows to enhance sampling...
> 

I don't know if I can make a convincing argument here, but intuitively, these 
windows would be sampling in a different ensemble, so the free energy landscape 
in these windows would be discontinuous with any adjacent windows that are done 
at different temperatures, and perhaps the forces required to restrain your 
peptide at a given COM distance will still result in a discontinuous PMF.  I 
would also suspect that g_wham can't handle this situation; it has a -temp flag, 
but it only takes one value.  So if you construct your PMF curve using WHAM, but 
supply incorrect or inconsistent information, I certainly wouldn't believe the 
result.

I guess the main point is, there are tons of published demonstrations of 
peptides and other molecules crossing a membrane with SMD and umbrella sampling, 
so it should be possible to generate stable configurations without any funny tricks.

-Justin

> best regards,
> Jianguo
> 
> 
> 
> ------------------------------------------------------------------------
> *From:* Justin A. Lemkul <jalemkul at vt.edu>
> *To:* Discussion list for GROMACS users <gmx-users at gromacs.org>
> *Sent:* Tuesday, 22 February 2011 09:35:37
> *Subject:* Re: [gmx-users] Can g_wham support using different 
> temperature for different windows?
> 
> 
> 
> Jianguo Li wrote:
>  > Dear all,
>  >
>  > I want to get the PMF of my peptide across the membrane bilayer. 
> First I pulled my peptide across the membrane and then did windowed 
> umbrella sampling along the reaction coordinates which is the z-distance 
> between peptide and membrane. However, I found that sampling is not 
> sufficient in some windows(e.g., around the center of the membrane). To 
> enhance the sampling, I am thinking to run the simulation in those 
> windows at higher temperature (e.g., 500K), but this will introduce a 
> bias. My question is: can g_wham remove the bias due to using different 
> temperatures in different windows?
>  >
>  > If g_wham cannot deal with the bias due to using different T, I may 
> need to do REMD in those windows. But that will be very expensive 
> computationally. Anybody have an idea of enhancing sampling in those 
> windows?
>  >
>  > Btw, I am using Martini CG model.
>  >
>  > Any suggestions will be highly appreciated, thank you!
>  >
> 
> A more straightforward approach is to (1) add more sampling windows or 
> (2) increase the force constant in regions where there's poor sampling, 
> or perhaps both.
> 
> -Justin
> 
>  > Cheers,
>  > Jianguo
>  >
> 
> -- ========================================
> 
> Justin A. Lemkul
> Ph.D. Candidate
> ICTAS Doctoral Scholar
> MILES-IGERT Trainee
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu | (540) 231-9080
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
> 
> ========================================
> -- gmx-users mailing list    gmx-users at gromacs.org 
> <mailto:gmx-users at gromacs.org>
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at 
> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> Please don't post (un)subscribe requests to the list. Use the www 
> interface or send it to gmx-users-request at gromacs.org 
> <mailto:gmx-users-request at gromacs.org>.
> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================