Merge pull request #389 from csbrasnett/mods-fix

fixed modifcation parsing

bin/polyply

@@ -74,12 +74,12 @@ def main(): # pylint: disable=too-many-locals,too-many-statements
                            help='A linear sequence of residue names.')
     seq_group.add_argument('-seqf', dest='seq_file', type=Path,
                            help='A graph input file (JSON|TXT|FASTA|IG)')
-    dna_group = parser_gen_itp.add_argument_group('DNA specifc options')
+    dna_group = parser_gen_itp.add_argument_group('DNA specific options')
     dna_group.add_argument('-dsdna', dest='dsdna', action='store_true',
                            help='complement single sequence to dsDNA sequence')
     modifications_group = parser_gen_itp.add_argument_group('Modifications group')
-    modifications_group.add_argument('-mods', dest='mods', action='append',
-                                     default=[], type=lambda s: s.split(":"),
+    modifications_group.add_argument('-mods', dest='mods', nargs='+',
+                                     default=[], type=lambda s: [i.split(':') for i in s.split(" ")][0],
                                      help=('Add a modification to a residue. The format is '
                                            '<resname><resid>:modification_name, '
                                            'e.g. ASP1:N-ter')

polyply/data/martini3/modifications.ff

@@ -35,3 +35,72 @@ BB {"replace": {"atype": "P1", "charge": 0.0}}
 NCAP-ter
 [ atoms ]
 BB {"replace": {"atype": "P2", "charge": 0.0}}
+
+[ modification ]
+GLU-HE1
+[ atoms ]
+SC1 {"replace": {"atype": "P2", "charge": 0}}
+
+[ modification ]
+GLU-HE2
+[ atoms ]
+SC1 {"replace": {"atype": "P2", "charge": 0}}
+
+[ modification ]
+ASP-HD1
+[ atoms ]
+SC1 {"replace": {"atype": "P2", "charge": 0}}
+
+[ modification ]
+ASP-HD2
+[ atoms ]
+SC1 {"replace": {"atype": "P2", "charge": 0}}
+
+[ modification ]
+LYS-LSN
+[ atoms ]
+SC2 {"resname": "LYS", "replace": {"atype": "SN6d", "charge": 0}}
+
+[ modification ]
+LYS-HZ3
+[ atoms ]
+SC2 {"replace": {"resname": "LYS"}}
+
+[ modification ]
+HIS-HP
+[ atoms ]
+BB  {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC1 {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC2 {"resname": "HIS", "replace": {"atype": "TP1dq", "charge": "0.5", "resname": "HIS"}}
+SC3 {"resname": "HIS", "replace": {"atype": "TP1dq", "charge": "0.5", "resname": "HIS"}}
+;[ edges ]
+;BB SC1
+;SC1 SC2
+;SC1 SC3
+;SC2 SC3
+
+[ modification ]
+HIS-HE
+[ atoms ]
+BB  {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC1 {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC2 {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC3 {"resname": "HIS", "replace": {"resname": "HIS"}}
+;[ edges ]
+;BB SC1
+;SC1 SC2
+;SC1 SC3
+;SC2 SC3
+
+[ modification ]
+HIS-HD
+[ atoms ]
+BB  {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC1 {"resname": "HIS", "replace": {"resname": "HIS"}}
+SC2 {"resname": "HIS", "replace": {"resname": "HIS", "atype": "TN5a"}}
+SC3 {"resname": "HIS", "replace": {"resname": "HIS", "atype": "TN6a"}}
+;[ edges ]
+;BB SC1
+;SC1 SC2
+;SC1 SC3
+;SC2 SC3

polyply/src/apply_modifications.py

@@ -32,10 +32,9 @@ def _patch_protein_termini(meta_molecule, ter_mods=['N-ter', 'C-ter']):
         protein_termini.append(last_mod)
     else:
         # if only one mod in ter_mods, apply the mod to both start and end residue
-        LOGGER.info("Only one terminal modification specified. "
+        LOGGER.warning("Only one terminal modification specified. "
                     f"Will apply {ter_mods[0]} to both {meta_molecule.nodes[0]['resname']}1 and {last_resname}{max_resid}")
         protein_termini.append(({'resid': max_resid, 'resname': last_resname}, ter_mods[0]))
-
     return protein_termini
 
 
@@ -60,7 +59,7 @@ def apply_mod(meta_molecule, modifications):
     molecule = meta_molecule.molecule
 
     if not molecule.force_field.modifications:
-        LOGGER.warning('No modifications present in forcefield, none will be applied')
+        LOGGER.info('No modifications present in forcefield, none will be applied')
         return meta_molecule
 
     for target, desired_mod in modifications:
@@ -76,13 +75,13 @@ def apply_mod(meta_molecule, modifications):
 
         target_residue = meta_molecule.nodes[target_resid - 1]
         # takes care to skip all residues that come from an itp file
-        if not target_residue.get('from_itp', 'False'):
-            LOGGER.warning("meta_molecule has come from itp. Will not attempt to modify.")
+        if target_residue.get('from_itp'):
+            LOGGER.info("meta_molecule has come from itp. Will not attempt to modify.")
             continue
         # checks that the resname is a protein resname as defined above
         if not vermouth.molecule.attributes_match(target_residue,
                                               {'resname': vermouth.molecule.Choice(protein_resnames.split("|"))}):
-            LOGGER.warning("The resname of your target residue is not recognised a protein resname. "
+            LOGGER.info("The resname of your target residue is not recognised as a protein resname. "
                            "Will not attempt to modify.")
             continue
 
@@ -114,13 +113,10 @@ class ApplyModifications(Processor):
 
     """
     def __init__(self, meta_molecule, modifications=[]):
-        self.target_mods = []
+        self.target_mods = _patch_protein_termini(meta_molecule)
         for resspec, val in modifications:
             self.target_mods.append((parse_residue_spec(resspec), val))
-        if len(self.target_mods) == 0:
-            self.target_mods = _patch_protein_termini(meta_molecule)
 
     def run_molecule(self, meta_molecule):
-
         apply_mod(meta_molecule, self.target_mods)
         return meta_molecule

polyply/tests/test_apply_modifications.py

@@ -26,6 +26,7 @@
 import polyply.src.ff_parser_sub
 import networkx as nx
 from vermouth.molecule import Interaction
+from polyply.src.meta_molecule import MetaMolecule
 
 
 @pytest.mark.parametrize('input_mods, expected',(
@@ -37,8 +38,6 @@
         [({'resid': 1, 'resname': 'A'}, 'Zwitter'),
          ({'resid': 3, 'resname': 'A'}, 'Zwitter')]
         )
-
-
 ))
 def test_annotate_protein(example_meta_molecule, input_mods, expected):
     """"
@@ -75,20 +74,25 @@ def test_mods_in_ff(caplog, ff_files, expected):
     with expected:
         assert len(ff.modifications) > 0
 
-@pytest.mark.parametrize('input_itp, molname, expected',
+@pytest.mark.parametrize('input_itp, molname, expected, text',
                          (
     (
     'ALA5.itp',
     'pALA',
-    False),
+    False,
+    None),
     ('PEO.itp',
      'PEO',
-    True)
-                         ))
-def test_apply_mod(input_itp, molname, expected, caplog):
+    True,
+     ("The resname of your target residue"
+      " is not recognised as a protein resname."
+      " Will not attempt to modify."))
+))
+def test_apply_mod(input_itp, molname, expected, caplog, text):
     """
     test that modifications get applied correctly
     """
+    caplog.set_level(logging.INFO)
     #make the meta molecule from the itp and ff files
     file_name = TEST_DATA / "itp" / input_itp
 
@@ -107,7 +111,12 @@ def test_apply_mod(input_itp, molname, expected, caplog):
     apply_mod(meta_mol, termini)
 
     if expected:
-        assert any(rec.levelname == 'WARNING' for rec in caplog.records)
+        for record in caplog.records:
+            if record.message == text:
+                assert True
+                break
+        else:
+            assert False
 
     else:
         #for each mod applied, check that the mod atom and interactions have been changed correctly
@@ -134,25 +143,69 @@ def test_apply_mod(input_itp, molname, expected, caplog):
                                                        meta=interaction.meta)
                             assert _interaction in meta_mol.molecule.interactions[interaction_type]
 
-@pytest.mark.parametrize('modifications, expected',
-     (
-             (
-                 [['A1', 'N-ter']],
-                 True
-             ),
 
+@pytest.mark.parametrize('adding, expected',
+                         (
+    (True,
+    True),
+    (False,
+     False)
+                         ))
+def test_from_itp(caplog, adding, expected):
+
+    caplog.set_level(logging.INFO)
+    #make the meta molecule from the itp and ff files
+    file_name = TEST_DATA / "itp" / "ALA5.itp"
+
+    ff = vermouth.forcefield.ForceField(name='martini3')
+
+    ff_lines = []
+    for file in ['aminoacids.ff', 'modifications.ff']:
+        with open(TEST_DATA/ "ff" / file) as f:
+            ff_lines += f.readlines()
+    polyply.src.ff_parser_sub.read_ff(ff_lines, ff)
+
+    meta_mol = MetaMolecule.from_itp(ff, file_name, "pALA")
+
+    if adding:
+        for node in meta_mol.nodes:
+            meta_mol.nodes[node]['from_itp'] = 'True'
+
+    termini = _patch_protein_termini(meta_mol)
+    apply_mod(meta_mol, termini)
+
+    found = False
+    expected_msg = "meta_molecule has come from itp. Will not attempt to modify."
+    for record in caplog.records:
+        if record.message == expected_msg:
+            found = True
+            break
+        else:
+            continue
+
+    assert found == expected
+
+@pytest.mark.parametrize('modifications, expected, text',
+     (
              (
                  [],
-                 True
+                 True,
+                 "No modifications present in forcefield, none will be applied"
              ),
 
      ))
-def test_ApplyModifications(example_meta_molecule, caplog, modifications, expected):
+def test_ApplyModifications(example_meta_molecule, caplog, modifications, expected, text):
 
+    caplog.set_level(logging.INFO)
 
     ApplyModifications(modifications=modifications,
                        meta_molecule=example_meta_molecule).run_molecule(example_meta_molecule)
 
     if expected:
-        assert any(rec.levelname == 'WARNING' for rec in caplog.records)
-
+        for record in caplog.records:
+            if record.message == text:
+                assert True
+                break
+        else:
+            assert False
+        assert any(rec.levelname == 'INFO' for rec in caplog.records)