pdb2pqr does not remove atoms as expected

[CyberCatQ]

Dear developers,

I am writing python code with pdb2pqr package, and I found that if I created a biomolecule include nucleic acids (PDB ID: 3AU6), calling biomolecule.set_termini() will remove the atom P of residue DG (chain T, residue index 2), but the atoms OP1/OP2 were kept.

This may break the nucleic acid residue and cause more problems.

Python code

from pdb2pqr import io
from pdb2pqr.main import setup_molecule
pdb_file = '3AU6.pdb'
pdb_list = io.get_molecule(pdb_file)[0]
definition = io.get_definitions()
biomolecule, definition, ligand = setup_molecule(pdb_list, definition, None)
biomolecule.set_termini()

And the pdb file is like this:

...
HETATM 4625  C6  DDG P   7       0.318 -29.896  14.522  1.00 78.77           C  
HETATM 4626  O6  DDG P   7       1.470 -30.120  14.970  1.00 78.13           O  
HETATM 4627  N1  DDG P   7      -0.537 -29.076  15.136  1.00 79.20           N  
HETATM 4628  C2  DDG P   7      -1.771 -28.834  14.656  1.00 79.68           C  
HETATM 4629  N2  DDG P   7      -2.564 -27.995  15.371  1.00 80.87           N  
HETATM 4630  N3  DDG P   7      -2.262 -29.385  13.515  1.00 78.01           N  
HETATM 4631  C4  DDG P   7      -1.500 -30.217  12.808  1.00 77.69           C  
TER
ATOM   4632  OP1 DG  T   2      16.134 -30.844  17.533  1.00101.02           O  
ATOM   4633  OP2 DG  T   2      14.316 -29.326  16.552  1.00103.39           O  
ATOM   4634  O5' DG  T   2      14.745 -29.259  19.007  1.00101.37           O  
ATOM   4635  C5' DG  T   2      15.093 -28.069  19.788  1.00100.31           C  
ATOM   4636  C4' DG  T   2      14.016 -26.980  19.897  1.00 99.38           C  
ATOM   4637  O4' DG  T   2      13.901 -26.265  18.634  1.00 99.16           O  
ATOM   4638  C3' DG  T   2      12.588 -27.455  20.191  1.00 98.70           C  
ATOM   4639  O3' DG  T   2      11.709 -26.388  20.617  1.00 97.61           O  
...

This residue also looks broken in pymol:

Are atoms OP1/OP2 not being removed as expected?
Thanks very much!