[gmx-users] Re: High Temprature MD with scaled params

[Vitaly Chaban]

> Hi,
>
> I have scaled the non-bonded parameters(charges and epsilon) of tip3p atoms
> in ffnonbonded.itp file(charmm forcefield). They are scaled by a ratio of
> 3/5.

What's the idea behind this scaling please?


>The simulation is then run at 500K for the protein water system. The
> temperature exploded after first 20 steps. This does not happen if I run the
> simulations with the unscaled parameters. I have also tried this with T=400
> and temperature remains stable.

Do you run NPT or NVT here? I believe the decrease of time-step can
solve any problem with a poorly equilibrated system.


Vitaly Chaban, Ph.D.