[gmx-users] Timestep in GROMACS simulation

[John Whittaker]

Hi,

> I'm still not clear what "sampling" means. Does it mean collecting data
> from the simulation?

Pretty much. The point of a molecular dynamics simulation is to calculate
structural, thermodynamic, etc... properties of some system from the
microscopic dynamics of atoms/molecules/whatever. Collecting this
information from the simulation is "sampling" that property from its
overall distribution. Typically, the longer you simulate, the better your
sampling is, but this isn't *always* true or feasible (e.g., development
of enhanced sampling methods to speed this process up).

By the way, the majority of the questions you have asked could probably be
answered much quicker and more in-depth if you just searched for them on
Google or took a look in the GROMACS manual. A massive amount of effort
has been put into tackling the MD "sampling problem", so you'll find lots
of information immediately.

Best,

John