Recognize that tleap reorders water residues BEFORE converting to GROMACS #67
Description
Right now, it takes the conversion to GROMACS to recognize that tleap always reorders waters to the end of the system, which means that internal data structures describing components are out of order until we do the GROMACS conversion. This should be fixed.
See https://github.com/MobleyLab/SolvationToolkit/blob/master/solvationtoolkit/solvated_mixtures.py#L484
It doesn't cause any bugs in AMBER setups that I am aware of, other than that the residues are not in the order expected (but the results should be correct).