[gmx-users] umbrella sampling tutorial

[Justin Lemkul]

On 3/15/15 9:57 PM, Ming Tang wrote:
> Dear Justin,
>
> I have tried many cases, and only your code runs smoothly. If I set pull-ngroups = 1, it will warn that pull-group2-name= Chain_A has not been defined.
> The code you gave me last time runs well. Though the COM between Chain_A and Chain_B remains the same over time, I found that Chain_A does dissociate from the structure and some particles fly away.

It's the same, but things move apart?  Sorry, this doesn't make sense to me.  Do 
the .mdp settings I provided before recover the behavior of the SMD portion of 
the tutorial or not?

> I am trying from the beginning once again and hope to find the reason. Besides, I got another problem. I want to probe the mechanical properties of a triple helix.
> I tried to fix one side (End) and pull the other side using the umbrella method:
>
> ; freeze groups
> freezegrps      = End
> freezedim       = Y Y Y
>
> ; COM pulling
> pull                = umbrella
> pull-geometry       = distance
> pull-dim            = Y N N
> pull-start          = yes
> pull-ngroups        = 2
> pull-ncoords        = 1
> pull-coord1-groups  = 1 2
> pull-group1-name    = End
> pull-group2-name    = pull
> pull-coord1-rate    = 0.005
> pull-coord1-k       = 1000
>
> WARNING: Listed nonbonded interaction between particles 179 and 190 at distance inf which is larger than the table limit 2.437 nm.
>
> This is likely either a 1,4 interaction, or a listed interaction inside
> a smaller molecule you are decoupling during a free energy calculation.
> Since interactions at distances beyond the table cannot be computed,
> they are skipped until they are inside the table limit again. You will
> only see this message once, even if it occurs for several interactions.
>
> IMPORTANT: This should not happen in a stable simulation, so there is
> probably something wrong with your system. Only change the table-extension
> distance in the mdp file if you are really sure that is the reason.
>

I'd venture a guess here that your system is probably blowing up because you're 
tugging on one molecule, which reaches a limit, distorts, and the simulation fails.

> If I just fix one of the three dimensions, it can run around 100 steps. Then, I turned to use SMD method to pull the triple helix along x direction. I tried the following method to generate the box, but got a fatal error.
> editconf -f complex.gro -o newbox.gro -c -d 1
> editconf -f complex.gro -o newbox.gro -center 5.63865 1.8837 1.9762 -box 20 3.8 4.0
>
> ; freeze groups
> freezegrps      = End
> freezedim       = Y Y Y
>
> ; COM pulling.
> pull                = umbrella
> pull-geometry       = direction-periodic
> pull-start          = yes
> pull-ngroups        = 2
> pull-ncoords        = 1
> pull-coord1-groups  = 1 2
> pull-group1-name    = End
> pull-group2-name    = pull
> pull-coord1-rate    = 0.01
> pull-coord1-k       = 1000
> pull-coord1-vec     = 1 0 0
>
> Fatal error:
> Can not have dynamic box while using pull geometry 'direction-periodic' (dim x)
>
> Could you teach me how can I build a fixed box? And is my SMD code right?
>

This means you can't use pressure coupling in the direction along which you're 
pulling.  Use NVT or align the system such that you are pulling along z, in 
conjunction with semiisotropic coupling and zero compressibility along z.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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