[gmx-users] positional restraint

[Alok]

hello David and Xavier,
                         Thanks for all your responses to my previous
mails.I am sorry,I still have to bother you with
some more queries about the constraint.In the
process i would also like to answer few questions
u have asked in your responding mails.

What is the range of differences you observe ?? Is it for the restrained
part or the rest of the protein ?

The Observed distance difference is for the restrained part (the atoms
from the constrained residue).I have not checked the distance difference
between the other atoms of the protein as i expect it to vary anyway since
i am constraining only one residue in my protein.

I present below data of distances (in Angstrom) between atoms of my
constained residue for the crystal structure as well as for the output
after Energy minimization using different Force constant.


                                     (Force Constant)
         Crystal      1000            10000              100000
 O-OD1    2.58         3.17            2.87               2.66
 O-HD21   4.86         5.27            5.13               4.95
 O-HD22   4.75         5.35            5.06               4.83
 H-OD1    4.73         4.76            4.75               4.75
 N-HD21   4.62         4.57            4.74               4.71
 N-HD22   5.58         5.62            5.65               5.62
 H-ND2    4.94         4.66            4.80               4.87

As we can observe even after value of 1000000,the atoms have not freezed
as u have mentioned in your mail.

my corresponding ".mdp " file for EM is as follows:

cpp                 =  /lib/cpp
define              =  -DFLEXIBLE
constraints         =  none
integrator          =  steep
nsteps              =  10000
;
;    Energy minimizing stuff
;
emtol               =  200
emstep              =  0.01
nstcgsteep          =  1000

nstcomm             =  1
ns_type             =  grid
rlist               =  1.0
rcoulomb            =  2.0
rvdw                =  1.2
Tcoupl              =  no
Pcoupl              =  no
gen_vel             =  no

If I am not able to freeze the atoms even after such high Force Constant,
then i am going wrong some where as you have suggested.Can you please
suggest the possible reason for the problem and would be thankfull if you
can shed some light into the possible solution.

                                          Thank you,
                                           Alok Jain



> David van der Spoel wrote:
>
>>On Mon, 2004-11-29 at 15:21 +0100, Xavier Periole wrote:
>>
>>
>>>Alok wrote:
>>>
>>>
>>>>                              Though i have increased the force
>>>>constant value,I am not sure about the detrimental structural effect (
>>>>I have read somewher that too large the force constant may result in
>>>>structural artifacts) it may have on my protein.So I would like to
>>>>know about the ideal value for the force constant such that the atoms
>>>>are constrained considerably and also that the protein does not
>>>>experience any unwanted structural deformation.
>>>>
>>>>
>>>It seems that there is definitely something wrong in your restrains. A
>>>force constant of 100000 should freeze your atoms
>>>
>>>
>>No! Only if you take a ridiculously small time step, otherwise they will
>>oscillate into oblivion.
>>
>>
> sorry for expressing myself a bit too fast. Oblivion seems top be much
> more adapted.
>
>>
>>
>>>at their original position. If you restrain only a part of your molecule
>>>it is reasonable to imagine that the rest of the protein
>>>is responding to "existing" forces.
>>>A force constant of about 500 to 1000 is ok !
>>>XAvier
>>>
>>>
>>>
>>>
>
>
> --
> ----------------------------------------------
>
>    Xavier Periole - Ph.D.
>
>    Dept. of Biophysical Chemistry / MD Group
>    Univ. of Groningen
>    Nijenborgh 4
>    9747 AG Groningen
>    The Netherlands
>
>    Tel: +31-503634329
>    Fax: +31-503634800
>    email: x.periole at chem.rug.nl
>    web-page: http://md.chem.rug.nl/~periole
>
> ----------------------------------------------
>
>
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
>