[gmx-users] restraint the CM of a molecules in any one of the directions.

[Justin Lemkul]

On 11/26/14 6:32 AM, Sridhar Kumar Kannam wrote:
> Thanks Justin,
> I agree its not a basic need, but an easy to implement (for the developers)
> and useful feature.
>
> I have added the below code to my mdp file.
>
> pull                      =  umbrella
> pull_geometry   =  distance
> pull_dim             =  Y Y N
> pull_start            =  yes
> pull_ngroups     =  1
> pull_group0       =  Protein
> pull_group1       =  Protein
> pull_init1            =  0.0
> pull_rate1          =  0.0
> pull_k1               =  1000
>
> Here, I am trying to constrain the CM of protein in x and y directions. The
> protein should be free to move in the z directions and rotate.
>
> The simulation is running fine without the pull code included but including
> the pull code resulting terminating the simulation with out any warnings or
> error messages. Is my code correct or am I missing anything ?
>

You can't have the same group serve as both reference (group0) and restrained 
(group1) group.  I'm surprised grompp didn't totally fail.  If you want an 
absolute restraint (not constraint, since it's only a bias) then pull_group0 
should be empty and pull_group1 would be the protein.  Heed the resulting grompp 
warnings about artifacts.

-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

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