What are the things I should keep in mind before doing an analysis of my gromacs simulation?

Question

I did a liquid argon simulation at 100k. I forgot during analysis that I need to accommodate for the Periodic Boundary Conditions before doing any analysis which included distance. What are other such general things I should keep in mind during my analysis ?

I am rather novice to computational Science and might have missed even the most trivial things.

Answer 1

I think that this question is too generic for a complete answer, as the latter would depend entirely on what you are simulating and what observables you are interested in. The only things that come to mind are

• Remember to use relative velocities (relative to the system's centre of mass) when computing dynamical observables (such as the mean-squared displacement or the intermediate scattering function).
• The radial distribution function makes sense only up to $r = L/2$, where $L$ is the side of the box.
• If the system crystallises, be aware that you probably need NPT (that is, constant-pressure) simulations to get meaningful results.
• Very close to a critical point (or to some other transitions/crossovers such as the percolation line) some quantities become heavily size-dependent, meaning that you need to do some extrapolations with the system size in order to get the "true" values.