[gmx-users] Bin size for Umbrella Sampling Analysis

[Kevin C Chan]

Dear Users,

I'm currently doing analysis of results of umbrella sampling. To guarantee
the reliability of the PMF profile, I plot the histograms to check the
overlapping of them. While we have to set a bin size for the analysis, I
found the bin size affect the overlapping of the histograms. For instance,
a larger bin size (e.g. 1600 for a PMF profile over 30A distance) will
cause a visually poorer overlapping of the histograms than a smaller bin
size (e.g. 200).

1) Is this normal? Can anyone share any experience on this?

2) What is the best bin size for an umbrella sampling analysis?