[gmx-users] LINCS crushes

[Mark Abraham]

On 1/11/2011 3:43 AM, Yuri Garmay wrote:
>
>
>     Try minimizing again now with all bonds constrained using the
>     output of the EM that ran.  Generally, if EM crashes, your system
>     contains some unresolvable clash or inappropriate geometry.
>      Perhaps you have now relaxed the bad interactions sufficiently to
>     proceed.
>
>
> I had examined structure before first minimization and if it was cause 
> I would have been wondered. However I have tried that advice, but this 
> didn't solve the problem. Minimization didn't crush, but didn't lead 
> to any significant atom position changes too.

It often doesn't.

> NPT position restrained simulation with all bonds constrained crushes 
> with minimized structure, but the same with hbonds constrained doesn't.

Jumping straight to NPT is sometimes a bad idea. See 
http://www.gromacs.org/Documentation/How-tos/Steps_to_Perform_a_Simulation

> The former produces exploded DMSO structures, the 
> latter produces visually correct trajectory. So, It seems that 
> structure is not the matter.

Your symptoms are consistent with a bad starting structure, wrong 
topology or invalid preparation protocol. Nothing can be ruled out yet.

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20111101/2eef3660/attachment.html>