mirror of
https://github.com/google-deepmind/alphafold3.git
synced 2026-06-02 11:54:36 +08:00
Fix some Structure type issues uncovered by Pyrefly
PiperOrigin-RevId: 891687351 Change-Id: I9c69179100e9328113a578d45e5f997b07d300a2
This commit is contained in:
Augustin Zidek
committed by
Copybara-Service
Copybara-Service
parent
0f82e0c94c
commit
58d22e2db5
@@ -108,59 +108,59 @@ class Bonds(table.Table):
|
||||
auth_seq_id: A (num_atom,) array of auth_seq_id strings.
|
||||
insertion_code: A (num_atom,) array of insertion code strings.
|
||||
"""
|
||||
mmcif_dict = collections.defaultdict(list)
|
||||
ptnr1_indices, ptnr2_indices = self.get_atom_indices(atom_key)
|
||||
cif = collections.defaultdict(list)
|
||||
p1_indices, p2_indices = self.get_atom_indices(atom_key)
|
||||
|
||||
mmcif_dict['_struct_conn.ptnr1_label_asym_id'] = chain_id[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_label_asym_id'] = chain_id[ptnr2_indices]
|
||||
mmcif_dict['_struct_conn.ptnr1_label_comp_id'] = res_name[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_label_comp_id'] = res_name[ptnr2_indices]
|
||||
mmcif_dict['_struct_conn.ptnr1_label_seq_id'] = res_id[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_label_seq_id'] = res_id[ptnr2_indices]
|
||||
mmcif_dict['_struct_conn.ptnr1_label_atom_id'] = atom_name[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_label_atom_id'] = atom_name[ptnr2_indices]
|
||||
cif['_struct_conn.ptnr1_label_asym_id'] = chain_id[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_label_asym_id'] = chain_id[p2_indices].tolist()
|
||||
cif['_struct_conn.ptnr1_label_comp_id'] = res_name[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_label_comp_id'] = res_name[p2_indices].tolist()
|
||||
cif['_struct_conn.ptnr1_label_seq_id'] = res_id[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_label_seq_id'] = res_id[p2_indices].tolist()
|
||||
cif['_struct_conn.ptnr1_label_atom_id'] = atom_name[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_label_atom_id'] = atom_name[p2_indices].tolist()
|
||||
|
||||
mmcif_dict['_struct_conn.ptnr1_auth_asym_id'] = auth_asym_id[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_auth_asym_id'] = auth_asym_id[ptnr2_indices]
|
||||
mmcif_dict['_struct_conn.ptnr1_auth_seq_id'] = auth_seq_id[ptnr1_indices]
|
||||
mmcif_dict['_struct_conn.ptnr2_auth_seq_id'] = auth_seq_id[ptnr2_indices]
|
||||
mmcif_dict['_struct_conn.pdbx_ptnr1_PDB_ins_code'] = insertion_code[
|
||||
ptnr1_indices
|
||||
]
|
||||
mmcif_dict['_struct_conn.pdbx_ptnr2_PDB_ins_code'] = insertion_code[
|
||||
ptnr2_indices
|
||||
]
|
||||
cif['_struct_conn.ptnr1_auth_asym_id'] = auth_asym_id[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_auth_asym_id'] = auth_asym_id[p2_indices].tolist()
|
||||
cif['_struct_conn.ptnr1_auth_seq_id'] = auth_seq_id[p1_indices].tolist()
|
||||
cif['_struct_conn.ptnr2_auth_seq_id'] = auth_seq_id[p2_indices].tolist()
|
||||
cif['_struct_conn.pdbx_ptnr1_PDB_ins_code'] = insertion_code[
|
||||
p1_indices
|
||||
].tolist()
|
||||
cif['_struct_conn.pdbx_ptnr2_PDB_ins_code'] = insertion_code[
|
||||
p2_indices
|
||||
].tolist()
|
||||
|
||||
label_alt_id = ['?'] * self.size
|
||||
mmcif_dict['_struct_conn.pdbx_ptnr1_label_alt_id'] = label_alt_id
|
||||
mmcif_dict['_struct_conn.pdbx_ptnr2_label_alt_id'] = label_alt_id
|
||||
cif['_struct_conn.pdbx_ptnr1_label_alt_id'] = label_alt_id
|
||||
cif['_struct_conn.pdbx_ptnr2_label_alt_id'] = label_alt_id
|
||||
|
||||
# We need to set this to make visualisation work in NGL/PyMOL.
|
||||
mmcif_dict['_struct_conn.pdbx_value_order'] = ['?'] * self.size
|
||||
cif['_struct_conn.pdbx_value_order'] = ['?'] * self.size
|
||||
|
||||
# We use a symmetry of 1_555 which is the no-op transformation. Other
|
||||
# values are used when bonds involve atoms that only exist after expanding
|
||||
# the bioassembly, but we don't support this kind of bond at the moment.
|
||||
symmetry = ['1_555'] * self.size
|
||||
mmcif_dict['_struct_conn.ptnr1_symmetry'] = symmetry
|
||||
mmcif_dict['_struct_conn.ptnr2_symmetry'] = symmetry
|
||||
cif['_struct_conn.ptnr1_symmetry'] = symmetry
|
||||
cif['_struct_conn.ptnr2_symmetry'] = symmetry
|
||||
bond_type_counter = collections.Counter()
|
||||
for bond_row in self.iterrows():
|
||||
bond_type = bond_row['type']
|
||||
bond_type_counter[bond_type] += 1
|
||||
mmcif_dict['_struct_conn.id'].append(
|
||||
cif['_struct_conn.id'].append(
|
||||
f'{bond_type}{bond_type_counter[bond_type]}'
|
||||
)
|
||||
mmcif_dict['_struct_conn.pdbx_role'].append(bond_row['role'])
|
||||
mmcif_dict['_struct_conn.conn_type_id'].append(bond_type)
|
||||
cif['_struct_conn.pdbx_role'].append(bond_row['role'])
|
||||
cif['_struct_conn.conn_type_id'].append(bond_type)
|
||||
|
||||
bond_types = np.unique(self.type)
|
||||
mmcif_dict['_struct_conn_type.id'] = bond_types
|
||||
cif['_struct_conn_type.id'] = bond_types
|
||||
unknown = ['?'] * len(bond_types)
|
||||
mmcif_dict['_struct_conn_type.criteria'] = unknown
|
||||
mmcif_dict['_struct_conn_type.reference'] = unknown
|
||||
cif['_struct_conn_type.criteria'] = unknown
|
||||
cif['_struct_conn_type.reference'] = unknown
|
||||
|
||||
return dict(mmcif_dict)
|
||||
return dict(cif)
|
||||
|
||||
|
||||
def concat_with_atom_keys(
|
||||
|
||||
@@ -50,6 +50,7 @@ class _UnsetSentinel(enum.Enum):
|
||||
UNSET = object()
|
||||
|
||||
|
||||
_UnsetType = Literal[_UnsetSentinel.UNSET]
|
||||
_UNSET = _UnsetSentinel.UNSET
|
||||
|
||||
|
||||
@@ -793,7 +794,7 @@ class Structure(table.Database):
|
||||
yield row | current_chain
|
||||
|
||||
def _iter_atom_ranges(
|
||||
self, boundaries: Sequence[int]
|
||||
self, boundaries: Sequence[int] | np.ndarray
|
||||
) -> Iterator[tuple[int, int]]:
|
||||
"""Iterator for (start, end) pairs from an array of start indices."""
|
||||
yield from itertools.pairwise(boundaries)
|
||||
@@ -803,7 +804,7 @@ class Structure(table.Database):
|
||||
|
||||
def _iter_residue_ranges(
|
||||
self,
|
||||
boundaries: Sequence[int],
|
||||
boundaries: Sequence[int] | np.ndarray,
|
||||
*,
|
||||
count_unresolved: bool,
|
||||
) -> Iterator[tuple[int, int]]:
|
||||
@@ -1131,20 +1132,20 @@ class Structure(table.Database):
|
||||
def copy_and_update(
|
||||
self,
|
||||
*,
|
||||
name: str | Literal[_UNSET] = _UNSET,
|
||||
release_date: datetime.date | None | Literal[_UNSET] = _UNSET,
|
||||
resolution: float | None | Literal[_UNSET] = _UNSET,
|
||||
structure_method: str | None | Literal[_UNSET] = _UNSET,
|
||||
name: str | _UnsetType = _UNSET,
|
||||
release_date: datetime.date | None | _UnsetType = _UNSET,
|
||||
resolution: float | None | _UnsetType = _UNSET,
|
||||
structure_method: str | None | _UnsetType = _UNSET,
|
||||
bioassembly_data: (
|
||||
bioassemblies.BioassemblyData | None | Literal[_UNSET]
|
||||
bioassemblies.BioassemblyData | None | _UnsetType
|
||||
) = _UNSET,
|
||||
chemical_components_data: (
|
||||
struc_chem_comps.ChemicalComponentsData | None | Literal[_UNSET]
|
||||
struc_chem_comps.ChemicalComponentsData | None | _UnsetType
|
||||
) = _UNSET,
|
||||
chains: structure_tables.Chains | None | Literal[_UNSET] = _UNSET,
|
||||
residues: structure_tables.Residues | None | Literal[_UNSET] = _UNSET,
|
||||
atoms: structure_tables.Atoms | None | Literal[_UNSET] = _UNSET,
|
||||
bonds: structure_tables.Bonds | None | Literal[_UNSET] = _UNSET,
|
||||
chains: structure_tables.Chains | None | _UnsetType = _UNSET,
|
||||
residues: structure_tables.Residues | None | _UnsetType = _UNSET,
|
||||
atoms: structure_tables.Atoms | None | _UnsetType = _UNSET,
|
||||
bonds: structure_tables.Bonds | None | _UnsetType = _UNSET,
|
||||
skip_validation: bool = False,
|
||||
) -> Self:
|
||||
"""Performs a shallow copy but with specified fields updated."""
|
||||
@@ -1322,15 +1323,15 @@ class Structure(table.Database):
|
||||
def copy_and_update_globals(
|
||||
self,
|
||||
*,
|
||||
name: str | Literal[_UNSET] = _UNSET,
|
||||
release_date: datetime.date | Literal[_UNSET] | None = _UNSET,
|
||||
resolution: float | Literal[_UNSET] | None = _UNSET,
|
||||
structure_method: str | Literal[_UNSET] | None = _UNSET,
|
||||
name: str | _UnsetType = _UNSET,
|
||||
release_date: datetime.date | _UnsetType | None = _UNSET,
|
||||
resolution: float | _UnsetType | None = _UNSET,
|
||||
structure_method: str | _UnsetType | None = _UNSET,
|
||||
bioassembly_data: (
|
||||
bioassemblies.BioassemblyData | Literal[_UNSET] | None
|
||||
bioassemblies.BioassemblyData | _UnsetType | None
|
||||
) = _UNSET,
|
||||
chemical_components_data: (
|
||||
struc_chem_comps.ChemicalComponentsData | Literal[_UNSET] | None
|
||||
struc_chem_comps.ChemicalComponentsData | _UnsetType | None
|
||||
) = _UNSET,
|
||||
) -> Self:
|
||||
"""Returns a shallow copy with the global columns updated."""
|
||||
|
||||
@@ -417,7 +417,7 @@ def to_mmcif_atom_site_and_bonds_table(
|
||||
coords_decimal_places: int,
|
||||
) -> Mapping[str, Sequence[str]]:
|
||||
"""Returns raw _atom_site and _struct_conn mmCIF tables."""
|
||||
raw_mmcif = collections.defaultdict(list)
|
||||
raw_mmcif: dict[str, Sequence[str]] = {}
|
||||
# Use [value] * num wherever possible since it is about 10x faster than list
|
||||
# comprehension in such cases. Also use f-strings instead of str() - faster.
|
||||
total_atoms = atoms.size * atoms.num_models
|
||||
|
||||
Reference in New Issue
Block a user