Remove bond-order guessing.

Remove guess_bond_order function and associated tests.  Bond orders are
already specified in the MOL2 file and the existing code wasn't very
good at guessing them.

pdb2pqr/pdb2pqr/ligand/__init__.py

@@ -24,25 +24,6 @@ PARSE_RADII = {
     "F": 1.20, "P": 1.90, "Cl": 1.75}
 
 
-# TODO - this belongs in a configuration file somewhere other than here.
-# Bond lengths from
-# http://www.chem.swin.edu.au/modules/mod2/bondlen.html
-# We should get a better reference
-_BOND_LENGTH_DICTS = [
-    {"atom1": 'C', "atom2": 'C', "length": 1.54, "type": "single"},
-    {"atom1": 'C', "atom2": 'C', "length": 1.34, "type": "double"},
-    {"atom1": 'C', "atom2": 'C', "length": 1.20, "type": "triple"},
-    {"atom1": 'C', "atom2": 'C', "length": 1.40, "type": "aromatic"},
-    {"atom1": 'C', "atom2": 'O', "length": 1.43, "type": "single"},
-    {"atom1": 'C', "atom2": 'O', "length": 1.21, "type": "double"},
-    {"atom1": 'C', "atom2": 'N', "length": 1.47, "type": "single"},
-    {"atom1": 'C', "atom2": 'N', "length": 1.25, "type": "double"},
-    {"atom1": 'C', "atom2": 'N', "length": 1.16, "type": "triple"},
-    {"atom1": 'C', "atom2": 'N', "length": 1.34, "type": "aromatic"},
-    {"atom1": 'N', "atom2": 'N', "length": 1.35, "type": "aromatic"}
-]
-BOND_LENGTHS = pandas.DataFrame(_BOND_LENGTH_DICTS)
-
 # Numbers of valence electrons for the groups of the periodic table
 VALENCE_BY_GROUP = {1: 1, 2: 2, 13: 3, 14: 4, 15: 5, 16: 6, 17: 7, 18: 8}
 # Groups of the periodic table

pdb2pqr/pdb2pqr/ligand/mol2.py

@@ -8,7 +8,7 @@ from collections import OrderedDict
 from itertools import combinations
 from numpy import array
 from numpy.linalg import norm
-from . import BOND_LENGTHS, VALENCE_BY_ELEMENT, NONBONDED_BY_TYPE
+from . import VALENCE_BY_ELEMENT, NONBONDED_BY_TYPE
 
 
 _LOGGER = logging.getLogger(__name__)
@@ -53,29 +53,6 @@ class Mol2Bond:
         fmt = "{b.atoms[0].name:s} {b.type:s}-bonded to {b.atoms[1].name:s}"
         return fmt.format(b=self)
 
-    @property
-    def guess_bond_order(self):
-        """Attempt to determine the order of this bond.
-
-        Return:
-            string with order of bond or None
-        """
-        _LOGGER.warning("Ignoring bond type: %s", self.type)
-        type1 = self.atoms[0].type.split(".")[0]
-        type2 = self.atoms[1].type.split(".")[0]
-        types = sorted(type1, type2)
-        bond_lengths = BOND_LENGTHS.loc[
-            (BOND_LENGTHS["atom1"] == types[0])
-            & (BOND_LENGTHS["atom2"] == types[1])]
-        best_type = None
-        best_fit = BOND_DIST
-        for _, row in bond_lengths.iterrows():
-            if abs(self.length - row["length"]) < best_fit:
-                best_fit = abs(self.length - row["length"])
-                best_type = row["type"]
-        return best_type
-
-
 class Mol2Atom:
     """MOL2 molecule atoms."""
     def __init__(self):

pdb2pqr/tests/ligand_results.py

@@ -69,27 +69,6 @@ RING_RESULTS = {
 }
 
 
-BOND_RESULTS = {
-    "cyclohexane.mol2": 6 * ["single"],
-    "ethanol.mol2": [
-        "single", "single", None, "single", "single"],
-    "glycerol.mol2": [
-        None, "single", "single", "single", "single", None, "single", None],
-    "acetylcholine.mol2": [
-        "single", "double", "single", "single", "single", "single", "single",
-        "single", "single"],
-    "acetonitrile.mol2": [
-        "triple", "single"],
-    "pyrrole.mol2": [
-        "aromatic", "aromatic", "aromatic", "aromatic", "aromatic", None],
-    "fatty-acid.mol2": [
-        "double", "double", "single", "single", "single", "single", "double",
-        "single", "single", "single", "single"],
-    "tetramethylammonium.mol2": ["single", None, "single", "single"],
-    "naphthalene.mol2": 11 * ["aromatic"]
-}
-
-
 FORMAL_CHARGE_RESULTS = {
     "1HPX-ligand.mol2": 87*[0],
     "1QBS-ligand.mol2": 80*[0],

pdb2pqr/tests/ligand_test.py

@@ -7,7 +7,7 @@ import pandas as pd
 from numpy.testing import assert_almost_equal
 from pdb2pqr.ligand import parameterize
 import common
-from ligand_results import TORSION_RESULTS, RING_RESULTS, BOND_RESULTS
+from ligand_results import TORSION_RESULTS, RING_RESULTS
 from ligand_results import FORMAL_CHARGE_RESULTS, PARTIAL_CHARGE_RESULTS
 
 
@@ -16,7 +16,7 @@ _LOGGER.warning("Need functional and regression test coverage for --ligand")
 _LOGGER.error("Still haven't figured out radii")
 
 
-ALL_LIGANDS = set(TORSION_RESULTS) | set(BOND_RESULTS) | set(RING_RESULTS)
+ALL_LIGANDS = set(TORSION_RESULTS) | set(RING_RESULTS)
 ALL_LIGANDS |= {
     "1HPX-ligand.mol2", "1QBS-ligand.mol2", "1US0-ligand.mol2", "adp.mol2",
     "acetate.mol2"}
@@ -154,26 +154,6 @@ def test_rings(input_mol2):
             _LOGGER.debug(str_)
 
 
-@pytest.mark.parametrize("input_mol2", ALL_LIGANDS)
-def test_bonds(input_mol2):
-    """Test detection of bond types."""
-    ligand = parameterize.ParameterizedMolecule()
-    mol2_path = Path("tests/data") / input_mol2
-    with open(mol2_path, "rt") as mol2_file:
-        ligand.read(mol2_file)
-    results = BOND_RESULTS[input_mol2]
-    for ibond, bond in enumerate(ligand.bonds):
-        try:
-            if bond.bond_order != results[ibond]:
-                err = "Incorrect order for %s. Got %s, expected %s" % (
-                    str(bond), bond.bond_order, results[ibond])
-                raise ValueError(err)
-        except IndexError:
-            err = "Add test for %s -- %s (%s)" % (
-                input_mol2, str(bond), bond.bond_order)
-            raise IndexError(err)
-
-
 @pytest.mark.parametrize("input_pdb", ["1HPX", "1QBS", "1US0"], ids=str)
 def test_ligand_protein(input_pdb, tmp_path):
     """PROPKA non-regression tests on proteins without ligands."""