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

[Justin Lemkul]

On 1/6/15 9:48 PM, Sridhar Kumar Kannam wrote:
> Hi Justin,
>
> Few weeks back I tried flat-bottom restraints also. It did not work as the inner
> and outer cylinder radius are equal and very small (to keep the protein close to
> the pore axis).
>
> Based on your tutorial on Umbrealla sampling I am using the below pull code.
> The nanopore (and hence its centre of mass) is fixed in my simulations.
>
> pull                               =  umbrella
> pull_geometry            =  distance
> pull_dim                      =  Y Y N
> pull_group1                =  Protein
> pull_group0                =  Nanopore
> pull_ngroups              =  1
> pull_rate1                   =  0.0
> pull_k1                        =  5000
>
> pull_init1                     =  0.0
> pull_start                    =  yes
>
> The reference CM co-inside with the pore axis, to which I want to constrain the
> protein. When I run the simulation, the protein is moving quite a far,  2 to 3
> nm from the axis.
> With the above pull code we expect the protein to stay close to the pore axis,
> i.e., can we restrain the CM distance between the two groups protein and
> nanopore (which in this case is zero distance) ?
>

Restraining the protein COM to be coincident with the pore COM is indeed the 
best way to do this; probably "distance" geometry is not the right way to go, 
though.  Try "position" (and other applicable modifications to other settings) 
instead.

-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

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