[gmx-users] How to determin the simulation time

[Mark Abraham]

On 11/09/2012 2:02 PM, xiaojing gong wrote:
> Dear all,
>          Typically, we use over 100 ns to simulate the transport progress.
> But *How the sufficient number of iterations and these times has been
> determined? Are there some convergence tests ?*

There ought to be, but in practice things vary a lot with how you do the 
simulation and the properties of the observable. Look at what others are 
doing for similar simulations.

Mark