[gmx-users] molecules blowing up

[X.Periole]

On Sun, 19 Feb 2006 09:22:41 -0800 (PST)
  Yegor Isakov <gromacs_problems at yahoo.com> wrote:
> I appreciate if  anybody could answer my question?
> ,
>       My molecules continue blowing out. I have got a 
>peptide where a  carbonyl group is connected to cysteine 
>through S atom (-S-CO-). I inserted my topology (the 
>whole set including peptide.top, peptide.gro, and 
>posre.itp ) into gmxdemo script (water) and run it the 
>same way as the demo cpeptide. My molecule blows out at 
>the third step of the run.  In the program output I have 
>got a message that the force is concentrated on the S 
>atom. 
>        I have got another molecule where carbonyl group 
>was attached to protein through NH group (-NH-CO-). This 
>bond is stable enough and the molecule runs fine up to 
>3000 steps.
> Thus, I concluded that the problem is that (-S-CO-) 
>bond. Actually I did not find  this bond in the libraries 
>and I introduced it myself into my topology by just 
>adding the the physical length of the bond. I believe 
>that's not enough. 
>      My question is: what I need to do to introduce into 
>my MD model a new bond that was not previously described 
>in those libraries? The manual does not cover this topic 
>explicitly. 
>
it is called distance restraints, 4.2.14 in manual gmx3.2.

XAvier