The documentation regarding the relative rmsd constraint (CONS RMSD RELATIVE) in c32a1 seems to be wrong in that, as far I can tell, the energy is derived from RMSD1-RMSD2 instead of RMSD2-RMSD1... Can anyone confirm this?

I believe RMSDs 1 & 2 are defined relative to the main & comp coordinate sets respectively when [COMP] is not used in the command line.

Many thanks,