Add support for openff-nagl for generation of partial charges

.github/workflows/devel.yaml

@@ -36,7 +36,7 @@ jobs:
       SIRE_DONT_PHONEHOME: 1
       SIRE_SILENT_PHONEHOME: 1
     steps:
-      - uses: conda-incubator/setup-miniconda@v2
+      - uses: conda-incubator/setup-miniconda@v3
         with:
           auto-update-conda: true
           python-version: ${{ matrix.python-version }}
@@ -49,7 +49,7 @@ jobs:
         run: git clone -b devel https://github.com/openbiosim/biosimspace
 #
       - name: Setup Conda
-        run: mamba install -y -c conda-forge boa anaconda-client packaging=21 pip-requirements-parser
+        run: mamba install -y -c conda-forge boa anaconda-client packaging pip-requirements-parser
 #
       - name: Update Conda recipe
         run: python ${{ github.workspace }}/biosimspace/actions/update_recipe.py

.github/workflows/main.yaml

@@ -30,7 +30,7 @@ jobs:
       SIRE_DONT_PHONEHOME: 1
       SIRE_SILENT_PHONEHOME: 1
     steps:
-      - uses: conda-incubator/setup-miniconda@v2
+      - uses: conda-incubator/setup-miniconda@v3
         with:
           auto-update-conda: true
           python-version: ${{ matrix.python-version }}
@@ -43,7 +43,7 @@ jobs:
         run: git clone -b main https://github.com/openbiosim/biosimspace
 #
       - name: Setup Conda
-        run: mamba install -y -c conda-forge boa anaconda-client packaging=21 pip-requirements-parser
+        run: mamba install -y -c conda-forge boa anaconda-client packaging pip-requirements-parser
 #
       - name: Update Conda recipe
         run: python ${{ github.workspace }}/biosimspace/actions/update_recipe.py

.github/workflows/pr.yaml

@@ -40,7 +40,7 @@ jobs:
       SIRE_SILENT_PHONEHOME: 1
       REPO: "${{ github.event.pull_request.head.repo.full_name || github.repository }}"
     steps:
-      - uses: conda-incubator/setup-miniconda@v2
+      - uses: conda-incubator/setup-miniconda@v3
         with:
           auto-update-conda: true
           python-version: ${{ matrix.python-version }}
@@ -53,7 +53,7 @@ jobs:
         run: git clone -b ${{ github.head_ref }} --single-branch https://github.com/${{ env.REPO }} biosimspace
 #
       - name: Setup Conda
-        run: mamba install -y -c conda-forge boa anaconda-client packaging=21 pip-requirements-parser
+        run: mamba install -y -c conda-forge boa anaconda-client packaging pip-requirements-parser
 #
       - name: Update Conda recipe
         run: python ${{ github.workspace }}/biosimspace/actions/update_recipe.py

python/BioSimSpace/IO/__init__.py

@@ -28,6 +28,9 @@
 .. autosummary::
     :toctree: generated/
 
+    clearCache
+    disableCache
+    enableCache
     fileFormats
     formatInfo
     readMolecules
@@ -38,3 +41,4 @@
 """
 
 from ._io import *
+from ._file_cache import *

python/BioSimSpace/IO/_file_cache.py

@@ -24,7 +24,7 @@
 __author__ = "Lester Hedges"
 __email__ = "[email protected]"
 
-__all__ = ["check_cache", "update_cache"]
+__all__ = ["clearCache", "disableCache", "enableCache"]
 
 import collections as _collections
 import hashlib as _hashlib
@@ -80,8 +80,43 @@ def __delitem__(self, key):
 # to the same format, allowing us to re-use the existing file.
 _cache = _FixedSizeOrderedDict()
 
+# Whether to use the cache.
+_use_cache = True
 
-def check_cache(
+
+def clearCache():
+    """
+    Clear the file cache.
+    """
+    global _cache
+    _cache = _FixedSizeOrderedDict()
+
+
+def disableCache():
+    """
+    Disable the file cache.
+    """
+    global _use_cache
+    _use_cache = False
+
+
+def enableCache():
+    """
+    Enable the file cache.
+    """
+    global _use_cache
+    _use_cache = True
+
+
+def _cache_active():
+    """
+    Internal helper function to check whether the cache is active.
+    """
+    global _use_cache
+    return _use_cache
+
+
+def _check_cache(
     system,
     format,
     filebase,
@@ -157,6 +192,8 @@ def check_cache(
     if not isinstance(skip_water, bool):
         raise TypeError("'skip_water' must be of type 'bool'.")
 
+    global _cache
+
     # Create the key.
     key = (
         system._sire_object.uid().toString(),
@@ -221,7 +258,7 @@ def check_cache(
         return ext
 
 
-def update_cache(
+def _update_cache(
     system,
     format,
     path,
@@ -284,6 +321,8 @@ def update_cache(
     if not isinstance(skip_water, bool):
         raise TypeError("'skip_water' must be of type 'bool'.")
 
+    global _cache
+
     # Convert to an absolute path.
     path = _os.path.abspath(path)
 

python/BioSimSpace/IO/_io.py

@@ -66,8 +66,9 @@
 from .._SireWrappers import System as _System
 from .. import _Utils
 
-from ._file_cache import check_cache as _check_cache
-from ._file_cache import update_cache as _update_cache
+from ._file_cache import _check_cache
+from ._file_cache import _update_cache
+from ._file_cache import _cache_active
 
 
 # Context manager for capturing stdout.
@@ -430,12 +431,9 @@ def readMolecules(
         )
         _has_gmx_warned = True
 
-    # Glob string to catch wildcards and convert to list.
+    # Convert a single string to a list.
     if isinstance(files, str):
-        if not files.startswith(("http", "www")):
-            files = _glob(files)
-        else:
-            files = [files]
+        files = [files]
 
     # Check that all arguments are of type 'str'.
     if isinstance(files, (list, tuple)):
@@ -449,6 +447,15 @@ def readMolecules(
     else:
         raise TypeError("'files' must be of type 'str', or a list of 'str' types.")
 
+    # Glob all files to catch wildcards.
+    new_files = []
+    for file in files:
+        if not file.startswith(("http", "www")):
+            new_files += _glob(file)
+        else:
+            new_files.append(file)
+    files = new_files
+
     # Validate the molecule unwrapping flag.
     if not isinstance(make_whole, bool):
         raise TypeError("'make_whole' must be of type 'bool'.")
@@ -741,14 +748,17 @@ def saveMolecules(
     # Save the system using each file format.
     for format in formats:
         # Copy an existing file if it exists in the cache.
-        ext = _check_cache(
-            system,
-            format,
-            filebase,
-            match_water=match_water,
-            property_map=property_map,
-            **kwargs,
-        )
+        if _cache_active():
+            ext = _check_cache(
+                system,
+                format,
+                filebase,
+                match_water=match_water,
+                property_map=property_map,
+                **kwargs,
+            )
+        else:
+            ext = None
         if ext:
             files.append(_os.path.abspath(filebase + ext))
             continue
@@ -835,7 +845,10 @@ def saveMolecules(
             files += file
 
             # If this is a new file, then add it to the cache.
-            _update_cache(system, format, file[0], match_water=match_water, **kwargs)
+            if _cache_active():
+                _update_cache(
+                    system, format, file[0], match_water=match_water, **kwargs
+                )
 
         except Exception as e:
             msg = "Failed to save system to format: '%s'" % format

python/BioSimSpace/Parameters/_Protocol/_openforcefield.py

@@ -73,15 +73,37 @@
 
 _openff = _try_import("openff")
 
+# Initialise the NAGL support flag.
+_has_nagl = False
+
 if _have_imported(_openff):
     from openff.interchange import Interchange as _Interchange
     from openff.toolkit.topology import Molecule as _OpenFFMolecule
-    from openff.toolkit.topology import Topology as _OpenFFTopology
     from openff.toolkit.typing.engines.smirnoff import ForceField as _Forcefield
+
+    try:
+        from openff.toolkit.utils.nagl_wrapper import (
+            NAGLToolkitWrapper as _NAGLToolkitWrapper,
+        )
+
+        _has_nagl = _NAGLToolkitWrapper.is_available()
+        from openff.nagl_models import get_models_by_type as _get_models_by_type
+
+        _models = _get_models_by_type("am1bcc")
+        try:
+            # Find the most recent AM1-BCC release candidate.
+            _nagl = _NAGLToolkitWrapper()
+            _nagl_model = sorted(
+                [str(model) for model in _models if "rc" in str(model)], reverse=True
+            )[0]
+        except:
+            _has_nagl = False
+        del _models
+    except:
+        _has_nagl = False
 else:
     _Interchange = _openff
     _OpenFFMolecule = _openff
-    _OpenFFTopology = _openff
     _Forcefield = _openff
 
 # Reset stderr.
@@ -105,7 +127,9 @@
 class OpenForceField(_protocol.Protocol):
     """A class for handling protocols for Open Force Field models."""
 
-    def __init__(self, forcefield, ensure_compatible=True, property_map={}):
+    def __init__(
+        self, forcefield, ensure_compatible=True, use_nagl=True, property_map={}
+    ):
         """
         Constructor.
 
@@ -123,6 +147,11 @@ def __init__(self, forcefield, ensure_compatible=True, property_map={}):
             original molecule, e.g. the original atom and residue names will be
             kept.
 
+        use_nagl : bool
+            Whether to use NAGL to compute AM1-BCC charges. If False, the default
+            is to use AmberTools via antechamber and sqm. (This option is only
+            used if NAGL is available.)
+
         property_map : dict
             A dictionary that maps system "properties" to their user defined
             values. This allows the user to refer to properties with their
@@ -136,6 +165,12 @@ def __init__(self, forcefield, ensure_compatible=True, property_map={}):
             property_map=property_map,
         )
 
+        if not isinstance(use_nagl, bool):
+            raise TypeError("'use_nagl' must be of type 'bool'")
+
+        # Set the NAGL flag.
+        self._use_nagl = use_nagl
+
         # Set the compatibility flags.
         self._tleap = False
         self._pdb2gmx = False
@@ -291,6 +326,23 @@ def run(self, molecule, work_dir=None, queue=None):
                 else:
                     raise _ThirdPartyError(msg) from None
 
+        # Apply AM1-BCC charges using NAGL.
+        if _has_nagl and self._use_nagl:
+            try:
+                _nagl.assign_partial_charges(
+                    off_molecule, partial_charge_method=_nagl_model
+                )
+            except Exception as e:
+                msg = "Failed to assign AM1-BCC charges using NAGL."
+                if _isVerbose():
+                    msg += ": " + getattr(e, "message", repr(e))
+                    raise _ThirdPartyError(msg) from e
+                else:
+                    raise _ThirdPartyError(msg) from None
+            charge_from_molecules = [off_molecule]
+        else:
+            charge_from_molecules = None
+
         # Extract the molecular topology.
         try:
             off_topology = off_molecule.to_topology()
@@ -317,7 +369,9 @@ def run(self, molecule, work_dir=None, queue=None):
         # Create an Interchange object.
         try:
             interchange = _Interchange.from_smirnoff(
-                force_field=forcefield, topology=off_topology
+                force_field=forcefield,
+                topology=off_topology,
+                charge_from_molecules=charge_from_molecules,
             )
         except Exception as e:
             msg = "Unable to create OpenFF Interchange object!"