Clarify gmx msd -mw and -mol options
gmx msd is unclear in documentation exactly how -mw (mass-weighting) and -mol (per molecule diffusion coefficients) work. At minimum we need better documentation, but I'd like to reconsider the default behavior, which is in some cases non-intuitive, or could even be interpreted as buggy.
-mw (default = yes) | -mol | result |
---|---|---|
no | no | Masses are not used |
yes | no | Contributions to the msd per group are mass-weighted |
no | yes | -mw is ignored, -o gives MW mol-COM based averages, -mol gives MW per-mol diffusion coefficient |
yes | yes | Sames as with -nomw |
Some points here:
- It's counterintuitive to me that without -mol, we default to mass-weighted MSDs. That is, even though particles are not necessarily related, their contribution to the average MSD for a given tau is mass-weighted. I don't think there's a physical reason for that
- The -nomw option is silently ignored with -mol. There may not be a ton of reasons one might want non-mass weighted centers to calculate COM for, but right now we're giving the illusion of option.
- In the docs, it's unclear that with -mol, the -o output is mol-COM based