[gmx-users] Re: Re: Bonds break while Minimising using distance restraints

[plmallip at mail.uh.edu]

plmallip at mail.uh.edu wrote:

Hi Justin, thanks for your response. You are right. I get the following message in the .job file

"Warning: 1-4 interaction between 6151 and 6160 at distance 1.387 which is larger than the 1-4 table size 1.000 nm
These are ignored for the rest of the simulation
This usually means your system is exploding"

The atoms I restrained are 1703 & 1712 (6151 & 6160 atoms are also from the same residues). The distance between them is 12 A (there are 14 missing residues in between). My aim is to use distance restraints, without building missing residues in between. Is there any way I can overcome this warning why minimizing the system?

Thanks & regards,
Latha.


> Dear colleagues,
>
> I need to use simple distance restraints of 12.5 A between two CA atoms
> of two residues. I am using the following lines in the .itp file
>
> #ifdef DDISRES
> [distance_restraints]
> ;ai   aj type index type' low   up1  up2  fac
> 1703 1712 1    0     1    11.5  12.0 12.5 1.0
>
> The first feww line of .mdp file for minimisation of protein alone is
>
> ; Preprocessing
> ;
> title               =  ${MOL}
> cpp                 =  /lib/cpp    ;Preprocessor
> define              = -DDISRES  ;For cg, and also steep
>

You are not actually applying your distance restraint.  If you have "#ifdef
DDISRES," then you would have to "define = -DDDISRES" in the .mdp file.  What
you probably meant to define was "#ifdef DISRES" in the topology.

>          After minimisation, the restarined residues & the adjacent
> bonds break. This results in fragmnets - residues alone, peptide bond
> alone and the rest of the protein. The distance restrained residues
> seems to try to move towards each other (6.04 A after minimisation) and
> this might have caused fragmentation. I tried to use various upper and
> lower values for bond length so as to increase flexibility. But, still I
> end up in the fragments.
>

Bonds don't break in classical MD, this is just an artifact of visualization,
probably from nasty steric clashes within your structure.


>         Some of the suggestions in the archive says VMD doesn't show
> bonds if they r above threshold value. When I checked the distances
> between the atoms, one of them is really long CO-CA bond 3.23 A
> (normally its 1.59A). This means the bond is no longer there.
>

No, the bond is there, VMD just isn't smart enough to see it :)  You are
probably well on your way to an explosion if you try to constrain bond lengths
with LINCS, however.

-Justin
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080903/8cccf3d8/attachment.html>