Merge pull request #321 from marrink-lab/box_from_coords

Box from coords

polyply/src/gen_coords.py

@@ -236,6 +236,22 @@ def gen_coords(toppath,
         LOGGER.info("loading grid",  type="step")
         grid = np.loadtxt(grid)
 
+    # where to get the box size from
+    if box is not None and not np.array_equal(topology.box, box):
+        msg = ("A box is provided via the -box command line "
+               "and the starting coordinates. We consider the "
+               "the box of starting coordinates as correct. ")
+        LOGGER.warning(msg,  type="warning")
+        box = topology.box
+    elif topology.box is not None:
+        box = topology.box
+        if density is not None:
+            msg = ("A density is provided via the command line, "
+                   "but the starting coordinates define a box."
+                   "Will try to pack all molecules in the box "
+                   "provided with starting coordinates.")
+            LOGGER.warning(msg,  type="warning")
+
     # do a sanity check
     LOGGER.info("checking residue integrity",  type="step")
     check_residue_equivalence(topology)

polyply/src/topology.py

@@ -46,8 +46,10 @@ def _coord_parser(path, extension):
             molecule.merge_molecule(new_mol)
     else:
         molecule = molecules
+
+    box = molecule.box
     positions = np.array(list(nx.get_node_attributes(molecule, "position").values()))
-    return positions
+    return positions, box
 
 
 def replace_defined_interaction(interaction, defines):
@@ -213,6 +215,8 @@ class Topology(System):
         A dictionary of all typed parameter
     defines: list
         A list of everything that is defined
+    box: np.array(3,1)
+        Box vectors as a, b, c in nanometers
     """
 
     def __init__(self, force_field, name=None):
@@ -228,6 +232,7 @@ def __init__(self, force_field, name=None):
         self.persistences = []
         self.distance_restraints = defaultdict(dict)
         self.volumes = {}
+        self.box = None
 
     def preprocess(self):
         """
@@ -425,7 +430,7 @@ def add_positions_from_file(self, path, skip_res=[], resolution='mol'):
         """
         path = Path(path)
         extension = path.suffix.casefold()[1:]
-        positions = _coord_parser(path, extension)
+        positions, self.box = _coord_parser(path, extension)
         max_coords = len(positions)
         total = 0
         for meta_mol in self.molecules:

polyply/tests/test_data/topology_test/meta.gro

@@ -2,4 +2,4 @@ PMMA test file
   2
     1PMMA    C1    1  -0.024  -0.125   0.104
     1PMMA    C2    2  -0.089  -0.100  -0.028
-  10.00000  10.00000  10.00000
+  10.00000  11.00000  12.00000

polyply/tests/test_data/topology_test/test.gro

@@ -14,4 +14,4 @@ PMMA test file
     2PMMA    O1   12   4.482   6.483   9.965
     2PMMA    O2   13   4.579   6.512   9.772
     2PMMA    C5   14   4.632   6.624   9.842
-  10.00000  10.00000  10.00000
+  10.00000  11.00000  12.00000

polyply/tests/test_data/topology_test/test.pdb

@@ -1,3 +1,4 @@
+CRYST1  118.249  120.688  110.944  90.00  90.00  90.00 P 1           1
 ATOM      1 BB   MET A   1      17.193  61.796  62.315  1.00  0.00              
 ATOM      2 SC1  MET A   1      20.670  62.642  62.148  1.00  0.00              
 ATOM      3 BB   TYR A   2      16.691  62.571  65.956  1.00  0.00              

polyply/tests/test_gen_coords_logic.py

@@ -47,6 +47,53 @@ def test_no_positions_generated(tmp_path, monkeypatch):
         assert np.all(molecule_out.nodes[node]['position'] ==
                       molecule_in.nodes[node]['position'])
 
+@pytest.mark.parametrize('box_input, box_ref, density, warning', [
+                        # box from input coordinates
+                        (None, np.array([11.0, 11.0, 11.0]), None, None),
+                        # box from input coordinates overwrites
+                        (np.array([5.0, 5.0, 5.0]), np.array([11.0, 11.0, 11.0]), None, "warn1"),
+                        # box from input coordinates and density from CLI
+                        (None, np.array([11.0, 11.0, 11.0]), 1000, "warn2"),
+                        ])
+def test_box_input(tmp_path, caplog, box_input, box_ref, density, warning):
+    """
+    Here we test that the correct box is chosen, in case there
+    are conflicting inputs.
+    """
+    warnings = {"warn1": ("A box is provided via the -box command line "
+                          "and the starting coordinates. We consider the "
+                          "the box of starting coordinates as correct. "),
+                "warn2": ("A density is provided via the command line, "
+                          "but the starting coordinates define a box."
+                          "Will try to pack all molecules in the box "
+                          "provided with starting coordinates."),}
+
+    top_file = TEST_DATA + "/topology_test/system.top"
+    pos_file = TEST_DATA + "/topology_test/complete.gro"
+    out_file = tmp_path / "out.gro"
+
+    with caplog.at_level(logging.WARNING):
+        gen_coords(toppath=top_file,
+                   coordpath=pos_file,
+                   outpath=out_file,
+                   name="test",
+                   box=box_input,
+                   density=density,)
+
+        molecule_out = read_gro(out_file, exclude=())
+        assert np.array_equal(molecule_out.box, box_ref)
+        if warning:
+            for record in caplog.records:
+                if record.levelname == "WARNING":
+                    assert str(record.msg) == warnings[warning]
+                    break
+            else:
+                assert False
+        else:
+            for record in caplog.records:
+                if record.levelname == "WARNING":
+                    assert False
+
 def test_backmap_only(tmp_path, monkeypatch):
     """
     Only meta_mol positions are defined so others have to be backmapped.

polyply/tests/test_topology.py

@@ -46,6 +46,10 @@ def test_add_positions_from_gro():
         """
         top = Topology.from_gmx_topfile(TEST_DATA + "/topology_test/system.top", "test")
         top.add_positions_from_file(TEST_DATA + "/topology_test/test.gro")
+
+        # check that the box is correctly read
+        assert np.allclose(top.box, np.array([10.0, 11.0, 12.0]))
+
         for node in top.molecules[0].molecule.nodes:
             if node < 14:
                 assert "position" in top.molecules[0].molecule.nodes[node].keys()
@@ -70,6 +74,10 @@ def test_add_meta_positions_from_gro():
         """
         top = Topology.from_gmx_topfile(TEST_DATA + "/topology_test/system.top", "test")
         top.add_positions_from_file(TEST_DATA + "/topology_test/meta.gro", resolution="meta_mol")
+
+        # check that the box is correctly read
+        assert np.allclose(top.box, np.array([10.0, 11.0, 12.0]))
+
         for node in top.molecules[0].nodes:
             if node != 2:
                 assert "position" in top.molecules[0].nodes[node].keys()
@@ -90,6 +98,9 @@ def test_add_positions_from_pdb():
         top = Topology.from_gmx_topfile(TEST_DATA + "/topology_test/pdb.top", "test")
         top.add_positions_from_file(TEST_DATA + "/topology_test/test.pdb")
 
+        # check that the box is correctly read
+        assert np.allclose(top.box, np.array([11.8249, 12.0688, 11.0944]))
+
         pdb_mols = read_pdb(TEST_DATA + "/topology_test/test.pdb")
         for idx, meta_mol in enumerate(top.molecules):
             for node in meta_mol.molecule.nodes:

setup.cfg

@@ -35,7 +35,7 @@ install-requires =  # ?? requires-dist?
     numpy
     decorator == 4.4.2
     networkx ~= 2.0
-    vermouth >= 0.9.1
+    vermouth >= 0.9.6
     scipy >= 1.6.0
     tqdm
 zip-safe = False