reduce nbr_radius to something reasonable for DNA params (#4933)
It would appear dna has had its NBR_RADIUS set to 1000 for a very long time.
This isn't super scientific (I just used the calculated radius of GUA from molfile_to_params.py) for all of them (+0.1)
Perhaps in the future it might be smart to remake them in-case something else is wrong, but frank says that he has some DNA fixes #4871 and that these params seem to work well for him in that branch so its likely that only the NBR_RADIUS needs to be updated
danpf 2020-08-26 NBR WAS set to 1000.0, I set it to the value molfile_to_params.py set for GUA on a single PDB + 0.1 it is possible this could be improved, but this seems good enough for now as it is no longer 1000.0.