Ordering problem for polymers in archive_to_universe

There is a mistake in the archive_to_universe() function in atom_utils.py: The monomers of a polymer are mis-indexed because they are stored non-consecutively in the system topology (atoms_group).