[gmx-users] what standard is successful when doing EM

[Mark Abraham]

On 4/02/2011 10:07 PM, gromacs wrote:
> Hi, guys
>
> Water molecule starting at atom 10578 can not be settled.
> Check for bad contacts and/or reduce the timestep.
>
> When i run 6x6x1.3nm water box. i used several methods to do EM. So i 
> can do several EM. But which .gro should i select as my input .gro for 
> my running£¿ Because when i use different .gro, i got different 
> results. So here i thought there should be a standard. Please note, 
> Even when the Max-force is below 50, there will also be different .gro 
> structure.

Yes, that's very normal. Potential energy surfaces of proteins are very 
rough. Small differences in EM algorithms find different minima or 
regions of minimas. Life's too short to be worried about finding the 
lowest one.

The usual purpose of EM is to relieve bad contacts before starting 
equilibration. As such, anything after EM that's rather better than what 
you started with is normally OK. So if equilibration succeeds, you 
probably did fine with EM. If not, do some detective work.

Mark