Merge branch 'main' into update-changelog-for-v040

.github/dependabot.yml

@@ -0,0 +1,7 @@
+version: 2
+updates:
+
+ - package-ecosystem: "github-actions"
+ directory: "/"
+ schedule:
+ interval: "daily"

.github/workflows/dev-ci.yaml

@@ -1,4 +1,4 @@
-name: Dev version CI
+name: Upstream nightly version CI
 on:
  schedule:
  # weekly tests, Sundays at midnight
@@ -24,14 +24,9 @@ jobs:
  matrix:
  os: [macOS-12, ubuntu-latest]
  python-version: ["3.10", "3.11", "3.12"]
- exclude:
- # mdtraj currently doesn't support MacOS py3.12
- - os: "macOS-latest"
- python-version: "3.12"
-
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Build information
  run: |

.github/workflows/examples-ci.yaml

@@ -39,7 +39,7 @@ jobs:
 
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Build information
  run: |

.github/workflows/gh-ci.yaml

@@ -39,9 +39,6 @@ jobs:
  exclude:
  - include-rdkit: false
  include-openeye: false
- # Can't support 3.12 on Mac yet
- - python-version: "3.12"
- os: "macOS-latest"
  # no openeye for 3.12 yet
  - include-openeye: true
  python-version: "3.12"
@@ -51,7 +48,7 @@ jobs:
 
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Build information
  run: |
@@ -111,7 +108,7 @@ jobs:
  python -m pytest -n 4 -v --cov=openff/nagl --cov-config=setup.cfg --cov-append --cov-report=xml --color=yes openff/nagl/
 
  - name: codecov
- uses: codecov/codecov-action@v3
+ uses: codecov/codecov-action@v4
  with:
  token: ${{ secrets.CODECOV_TOKEN }}
  file: ./coverage.xml
@@ -123,10 +120,10 @@ jobs:
  runs-on: ubuntu-latest
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Set up Python
- uses: actions/setup-python@v3
+ uses: actions/setup-python@v5
  with:
  python-version: "3.10"
 
@@ -147,10 +144,10 @@ jobs:
  runs-on: ubuntu-latest
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Set up Python
- uses: actions/setup-python@v3
+ uses: actions/setup-python@v5
  with:
  python-version: "3.10"
 
@@ -177,10 +174,10 @@ jobs:
  python-version: ["3.10", "3.11", "3.12"]
 
  steps:
- - uses: actions/checkout@v3
+ - uses: actions/checkout@v4
 
  - name: Install conda
- uses: conda-incubator/setup-miniconda@v2
+ uses: conda-incubator/setup-miniconda@v3
  with:
  python-version: ${{ matrix.python-version }}
  add-pip-as-python-dependency: true

devtools/conda-envs/docs_env.yaml

@@ -18,7 +18,7 @@ dependencies:
  - openff-toolkit-base >=0.11.1
  - openff-units
  - pydantic <2.0
- - rdkit !=2024.03.4
+ - rdkit
  - openeye-toolkits
 
  # database

devtools/conda-envs/examples_env.yaml

@@ -23,7 +23,7 @@ dependencies:
  - openff-units
  - openff-recharge
  - pydantic <3
- - rdkit !=2024.03.4
+ - rdkit
 
  # database
  - pyarrow

devtools/conda-envs/nightly.yaml

@@ -16,7 +16,7 @@ dependencies:
 
  # chemistry
  - pydantic <3
- - rdkit !=2024.03.4
+ - rdkit
  - scipy
 
  # database

devtools/conda-envs/test_cuda_env.yaml

@@ -22,7 +22,7 @@ dependencies:
  - openff-toolkit-base >=0.11.1
  - openff-units
  - pydantic <3
- - rdkit !=2024.03.4
+ - rdkit
  - openeye-toolkits
 
  # database

devtools/conda-envs/test_env_dgl_false.yaml

@@ -18,7 +18,7 @@ dependencies:
  - openff-toolkit-base >=0.11.1
  - openff-units
  - pydantic <3
- - rdkit !=2024.03.4
+ - rdkit
  - scipy
  - ambertools
 
@@ -37,4 +37,3 @@ dependencies:
  - pytest-cov
  - pytest-xdist
  - codecov
-

devtools/conda-envs/test_env_dgl_true.yaml

@@ -19,7 +19,7 @@ dependencies:
  - openff-toolkit-base >=0.11.1
  - openff-units
  - pydantic <3
- - rdkit !=2024.03.4
+ - rdkit
  - scipy
  - ambertools
 

docs/CHANGELOG.md

@@ -24,8 +24,8 @@ This version adds support for lookup tables.
 - [@j-wags]
 
 ### Added
-- Support for lookup tables and entries (PR #122)
-- Pins to DGL and RDKit for compatibility (PRs #117, #123)
+- Added lookup table support (`openff.nagl.lookups.AtomPropertiesLookupTable`) (PR #122)
+- Pins to DGL for compatibility (PR #117)
 
 ## v0.3.8 -- 2024-04-11
 

openff/nagl/lookups.py

@@ -0,0 +1,210 @@
+import types
+import typing
+
+import torch
+
+from openff.nagl._base.base import ImmutableModel
+from openff.nagl.utils._utils import is_iterable, potential_dict_to_list
+
+try:
+ from pydantic.v1 import Field, validator
+except ImportError:
+ from pydantic import Field, validator
+
+if typing.TYPE_CHECKING:
+ from openff.toolkit.topology import Molecule
+
+
+class PropertyProvenance(ImmutableModel):
+ """
+ Class for storing the provenance of a property
+ """
+ description: str = Field(
+ description="A description of the provenance"
+ )
+ versions: dict[str, str] = Field(
+ default_factory=dict,
+ description="The versions of the relevant software packages used to compute the property"
+ )
+
+class BasePropertiesLookupTableEntry(ImmutableModel):
+ inchi: str = Field(
+ description="The InChI of the molecule"
+ )
+ provenance: PropertyProvenance = Field(
+ description="The provenance of the property value"
+ )
+
+class AtomPropertiesLookupTableEntry(BasePropertiesLookupTableEntry):
+ """
+ Class for storing property lookup table entries
+ """
+ property_type: typing.Literal["atom"] = Field(
+ default="atom",
+ description="The type of the property"
+ )
+
+ mapped_smiles: str = Field(
+ description="The mapped SMILES of the molecule"
+ )
+
+ property_value: tuple[float, ...] = Field(
+ description=(
+ "The values of the property, ordered according to mapped SMILES"
+ )
+ )
+
+ def __len__(self):
+ return len(self.property_value)
+
+class BaseLookupTable(ImmutableModel):
+ """
+ Class for storing property lookup tables
+ """
+
+ property_name: str = Field(
+ description="The name of the property"
+ )
+
+
+class AtomPropertiesLookupTable(BaseLookupTable):
+ """
+ Class for storing property lookup tables for atom properties
+ """
+
+ property_type: typing.Literal["atom"] = Field(
+ default="atom",
+ description="The type of the property"
+ )
+
+ properties: types.MappingProxyType[str, AtomPropertiesLookupTableEntry] = Field(
+ description="The property lookup table"
+ )
+
+ @validator("properties", pre=True)
+ def _convert_property_lookup_table(cls, v):
+ """
+ Do two things:
+ 
+ 1. Account for an iterable being passed instead of a mapping
+ 2. Ignore the keys of the mapping and re-generate them from inchi
+ 
+ """
+ v = potential_dict_to_list(v)
+ if not is_iterable(v):
+ raise ValueError("The property lookup table must be an iterable")
+
+ if not all(isinstance(entry, AtomPropertiesLookupTableEntry) for entry in v):
+ raise ValueError("All entries must be AtomPropertiesLookupTableEntry instances")
+
+ return types.MappingProxyType({
+ entry.inchi: entry
+ for entry in v
+ })
+
+ def __len__(self) -> int:
+ return len(self.properties)
+
+ def __getitem__(self, key: str) -> AtomPropertiesLookupTableEntry:
+ return self.properties[key]
+
+ def __contains__(self, key: str) -> bool:
+ return key in self.properties
+
+ def lookup(self, molecule: "Molecule") -> torch.Tensor:
+ """
+ Look up the property value for a molecule
+
+ Parameters
+ ----------
+ molecule : openff.toolkit.topology.Molecule
+ The molecule to look up
+ 
+ Returns
+ -------
+ torch.Tensor
+ The property values, in the order of the molecule's atoms
+ 
+ Raises
+ ------
+ KeyError
+ If the property value cannot be found for this molecule
+ """
+ from openff.toolkit.topology import Molecule
+
+ inchi_key = molecule.to_inchi(fixed_hydrogens=True)
+ try:
+ entry = self.properties[inchi_key]
+ except KeyError:
+ raise KeyError(f"Could not find property value for molecule with InChI {inchi_key}")
+
+ assert len(entry) == molecule.n_atoms
+
+ # remap to query order
+ entry_molecule = Molecule.from_mapped_smiles(
+ entry.mapped_smiles,
+ allow_undefined_stereo=True
+ )
+
+ # first try with input bond orders and formal charges
+ is_isomorphic, query_to_entry_mapping = Molecule.are_isomorphic(
+ molecule,
+ entry_molecule,
+ return_atom_map=True,
+ )
+ if not is_isomorphic:
+ # try again without bond orders and formal charges.
+ # This should be enough to match the InChI
+
+ is_isomorphic, query_to_entry_mapping = Molecule.are_isomorphic(
+ molecule,
+ entry_molecule,
+ return_atom_map=True,
+ aromatic_matching=True,
+ formal_charge_matching=False,
+ bond_order_matching=False,
+ )
+
+ if not is_isomorphic:
+ # lastly -- skip stereochemistry matching
+ is_isomorphic, query_to_entry_mapping = Molecule.are_isomorphic(
+ molecule,
+ entry_molecule,
+ return_atom_map=True,
+ formal_charge_matching=False,
+ bond_order_matching=False,
+ # skip stereochemistry because matching inchi should be enough
+ atom_stereochemistry_matching=False,
+ bond_stereochemistry_matching=False,
+ )
+
+ assert is_isomorphic
+
+ # remap the property values to the query order
+ property_values = [
+ entry.property_value[query_to_entry_mapping[atom_index]]
+ for atom_index in range(molecule.n_atoms)
+ ]
+ return torch.tensor(property_values, dtype=torch.float32)
+
+
+
+
+
+LookupTableEntryType = typing.Union[AtomPropertiesLookupTableEntry]
+LookupTableType = typing.Union[AtomPropertiesLookupTable]
+
+LOOKUP_TABLE_CLASSES = {
+ "atom": AtomPropertiesLookupTable,
+}
+
+
+def _as_lookup_table(lookup_table_kwargs: dict) -> LookupTableType:
+ """
+ Convert a dictionary to a lookup table
+ """
+ if not isinstance(lookup_table_kwargs, BaseLookupTable):
+ lookup_table_type = lookup_table_kwargs["property_type"]
+ lookup_table_class = LOOKUP_TABLE_CLASSES[lookup_table_type]
+ return lookup_table_class(**lookup_table_kwargs)
+ return lookup_table_kwargs