[gmx-users] How many ps needs a simulation?

[Mark Abraham]

jahanshah ashkani wrote:
> Dear gromacs users,
> How many ps needs a simulation?

How long should a piece of string be?

It depends on what sort of physical insight you are hoping to achieve, 
how long the process you are studying takes to occur, how many degrees 
of freedom your system possess, how much computational time you're 
prepared to commit to it, how much data you're prepared to analyze, 
whether you can demonstrate convergence of some property of interest... 
For the record, though, referees are likely to laugh at simulations of 
biological systems which can be comfortably measured in picoseconds.

Mark