[gmx-users] Re: Pulling ion - US

[Thomas Schlesier]

I would also use the same residue from the pulling for the US.
One thing you should be aware of is the pulling dimension:
Now you have the pull-code only ativated for the z-direction. If you use 
this still in the US the ion can move freely in the xy-plane (freely in 
the sense of what is possible from the surrounding).

One extreme example:
The ion bounds to the protein and somehow (don't ask, think this wont 
happen in reality) and diffses away (in the xy-plane) after a long time 
it's so far away from the protein that the are no interaction with the 
protein and the ion interacts only with the surrounding water. Now you 
don't measure with the US potential the interaction of the ion with the 
protein, but the free diffusion of the ion.
This case wont happen, since the probability that the ion unbounds 
itself from the protein goes down to the cellar. But i hope you get the 
idea what the gerneral problem is. If you the pull-dim in each direction 
this problem wouldn't occur, since the movement of the ion is also 
restrained in the xy-plane.


Am 10.12.2012 21:40, schrieb gmx-users-request at gromacs.org:
> Would you also specify in each US window specific residue instead of
> the whole protein?
>
> Sreven
>
> On Mon, Dec 10, 2012 at 2:47 PM, Steven Neumann<s.neumann08 at gmail.com>  wrote:
>> >On Mon, Dec 10, 2012 at 2:11 PM, Justin Lemkul<jalemkul at vt.edu>  wrote:
>>> >>
>>> >>
>>> >>On 12/10/12 9:01 AM, Steven Neumann wrote:
>>>> >>>
>>>> >>>Dear Gmx Users,
>>>> >>>
>>>> >>>I am pulling away cation from the protein glutamic acid residue with:
>>>> >>>
>>>> >>>pull            = umbrella
>>>> >>>pull_geometry   = distance  ; simple distance increase
>>>> >>>pull_dim        = N N Y
>>>> >>>pull_start      = yes       ; define initial COM distance > 0
>>>> >>>pull_ngroups    = 1
>>>> >>>pull_group0     = Protein
>>>> >>>pull_group1     = NA
>>>> >>>pull_rate1      = 0.01
>>>> >>>pull_k1         = 500      ; kJ mol^-1  nm^-2
>>>> >>>
>>>> >>>I tried different pulling rates and simulation time to pull it 3 nm
>>>> >>>away. I tried pull rate of 0.1; 0.01 and 0.001. The interaction is so
>>>> >>>strong that the force reaches 600 kJ/mol/nm2 and they do not become
>>>> >>>separated - with position restraints protein looses its secondary
>>>> >>>structure and is draged by the ion - they do not become separated.
>>>> >>>
>>>> >>>Would you suggest constant force pulling in this case? Then I will
>>>> >>>extract initial coordinates for US windows. Can I use then US with
>>>> >>>harmonic potential in windows then and WHAM?
>>>> >>>
>>> >>
>>> >>You can generate coordinates in any way you wish.  I would think that,
>>> >>regardless of the pull method, setting pull_group0 to the actual residue to
>>> >>which the ion is coordinated would be significantly more effective than
>>> >>pulling with respect to the entire protein, though it seems rather strange
>>> >>that the dissociation of an ion would cause a protein to unfold.  A stronger
>>> >>force constant in pull_k1 may also help.
>>> >>
>>> >>-Justin
>> >
>> >Thank you Justin. That indeed helped.
>> >
>> >Steven