import pandas as pd
import numpy as np
from molsysmt._private.variables import is_all
from molsysmt._private.digestion import digest
from molsysmt.lib.series import occurrence_order
import string
[docs]
class Atoms_DataFrame(pd.DataFrame):
def __init__(self, n_atoms=0):
columns = ['atom_id', 'atom_name', 'atom_type', 'group_index', 'component_index', 'chain_index']
super().__init__(index=range(n_atoms), columns=columns)
self['atom_id'] = self['atom_id'].astype('Int64')
self['atom_name'] = self['atom_name'].astype(str)
self['atom_type'] = self['atom_type'].astype(str)
self['group_index'] = self['group_index'].astype('Int64')
self['component_index'] = self['component_index'].astype('Int64')
self['chain_index'] = self['chain_index'].astype('Int64')
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Groups_DataFrame(pd.DataFrame):
def __init__(self, n_groups=0):
columns = ['group_id', 'group_name', 'group_type', 'molecule_index']
super().__init__(index=range(n_groups), columns=columns)
self['group_id'] = self['group_id'].astype('Int64')
self['group_name'] = self['group_name'].astype(str)
self['group_type'] = self['group_type'].astype(str)
self['molecule_index'] = self['molecule_index'].astype('Int64')
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Molecules_DataFrame(pd.DataFrame):
def __init__(self, n_molecules=0):
columns = ['molecule_id', 'molecule_name', 'molecule_type', 'entity_index']
super().__init__(index=range(n_molecules), columns=columns)
self['molecule_id'] = self['molecule_id'].astype('Int64')
self['molecule_name'] = self['molecule_name'].astype(str)
self['molecule_type'] = self['molecule_type'].astype(str)
self['entity_index'] = self['entity_index'].astype('Int64')
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Entities_DataFrame(pd.DataFrame):
def __init__(self, n_entities=0):
columns = ['entity_id', 'entity_name', 'entity_type']
super().__init__(index=range(n_entities), columns=columns)
self['entity_id'] = self['entity_id'].astype('Int64')
self['entity_name'] = self['entity_name'].astype(str)
self['entity_type'] = self['entity_type'].astype(str)
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Components_DataFrame(pd.DataFrame):
def __init__(self, n_components=0):
columns = ['component_id', 'component_name', 'component_type']
super().__init__(index=range(n_components), columns=columns)
self['component_id'] = self['component_id'].astype('Int64')
self['component_name'] = self['component_name'].astype(str)
self['component_type'] = self['component_type'].astype(str)
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Chains_DataFrame(pd.DataFrame):
def __init__(self, n_chains=0):
columns = ['chain_id', 'chain_name', 'chain_type']
super().__init__(index=range(n_chains), columns=columns)
self['chain_id'] = self['chain_id'].astype('Int64')
self['chain_name'] = self['chain_name'].astype(str)
self['chain_type'] = self['chain_type'].astype(str)
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
[docs]
class Bonds_DataFrame(pd.DataFrame):
def __init__(self, n_bonds=0):
columns = ['atom1_index', 'atom2_index']
columns += ['order', 'type'] # extra columns -not necessary-
super().__init__(index=range(n_bonds), columns=columns)
self['atom1_index'] = self['atom1_index'].astype('Int64')
self['atom2_index'] = self['atom2_index'].astype('Int64')
self['order'] = self['order'].astype(str)
self['type'] = self['type'].astype(str)
def _reset(self, n_bonds=0):
columns = ['atom1_index', 'atom2_index']
columns += ['order', 'type'] # extra columns -not necessary-
super().__init__(index=range(n_bonds), columns=columns)
self['atom1_index'] = self['atom1_index'].astype('Int64')
self['atom2_index'] = self['atom2_index'].astype('Int64')
self['order'] = self['order'].astype(str)
self['type'] = self['type'].astype(str)
def _fix_null_values(self):
for column in self:
self[column]=self[column].fillna(pd.NA)
def _sort_bonds(self):
mask = self['atom1_index'] > self['atom2_index']
self.loc[mask, ['atom1_index', 'atom2_index']] = self.loc[mask, ['atom2_index', 'atom1_index']].values
self.sort_values(by=['atom1_index', 'atom2_index'], inplace=True)
self.reset_index(drop=True, inplace=True)
def _remove_empty_columns(self):
if (self['order']=='nan').all():
del self['order']
if (self['type']=='nan').all():
del self['type']
[docs]
class Topology():
@digest()
def __init__(self, n_atoms=0, n_groups=0, n_components=0, n_molecules=0, n_entities=0, n_chains=0, n_bonds=0,
skip_digestion=False):
self.reset_atoms(n_atoms=n_atoms)
self.reset_groups(n_groups=n_groups)
self.reset_components(n_components=n_components)
self.reset_molecules(n_molecules=n_molecules)
self.reset_entities(n_entities=n_entities)
self.reset_chains(n_chains=n_chains)
self.reset_bonds(n_bonds=n_bonds)
[docs]
def reset_atoms(self, n_atoms=0):
self.atoms = Atoms_DataFrame(n_atoms=n_atoms)
[docs]
def reset_groups(self, n_groups=0):
self.groups = Groups_DataFrame(n_groups=n_groups)
[docs]
def reset_components(self, n_components=0):
self.components = Components_DataFrame(n_components=n_components)
[docs]
def reset_molecules(self, n_molecules=0):
self.molecules = Molecules_DataFrame(n_molecules=n_molecules)
[docs]
def reset_entities(self, n_entities=0):
self.entities = Entities_DataFrame(n_entities=n_entities)
[docs]
def reset_chains(self, n_chains=0):
self.chains = Chains_DataFrame(n_chains=n_chains)
[docs]
def reset_bonds(self, n_bonds=0):
self.bonds = Bonds_DataFrame(n_bonds=n_bonds)
[docs]
@digest()
def remove(self, atom_indices=None, copy_if_None=False, skip_digestion=False):
if atom_indices is None:
if copy_if_None:
return self.copy()
else:
return self
else:
atom_indices_to_be_kept = np.setdiff1d(np.arange(self.n_atoms), atom_indices)
tmp_item = self.extract(atom_indices=atom_indices_to_be_kept, skip_digestion=True)
return tmp_item
[docs]
@digest(form='molsysmt.Topology')
def add(self, item, atom_indices='all', keep_ids=True, skip_digestion=False):
if is_all(atom_indices):
tmp_item = item.copy()
else:
tmp_item = item.extract(atom_indices=atom_indices, skip_digestion=True)
n_atoms = self.atoms.shape[0]
n_groups = self.groups.shape[0]
n_components = self.components.shape[0]
n_molecules = self.molecules.shape[0]
n_chains = self.chains.shape[0]
tmp_item.atoms['group_index'] += n_groups
tmp_item.atoms['component_index'] += n_components
tmp_item.atoms['chain_index'] += n_chains
tmp_item.groups['molecule_index'] += n_molecules
tmp_item.bonds['atom1_index'] += n_atoms
tmp_item.bonds['atom2_index'] += n_atoms
self.atoms = pd.concat([self.atoms, tmp_item.atoms], ignore_index=True, copy=False)
self.groups = pd.concat([self.groups, tmp_item.groups], ignore_index=True, copy=False)
self.molecules = pd.concat([self.molecules, tmp_item.molecules], ignore_index=True, copy=False)
self.components = pd.concat([self.components, tmp_item.components], ignore_index=True, copy=False)
self.chains = pd.concat([self.chains, tmp_item.chains], ignore_index=True, copy=False)
self.bonds = pd.concat([self.bonds, tmp_item.bonds], ignore_index=True, copy=False)
if not keep_ids:
self.rebuild_atoms(redefine_ids=True, redefine_types=False)
self.rebuild_groups(redefine_ids=True, redefine_types=False)
self.rebuild_components(redefine_indices=False, redefine_ids=(not keep_ids), redefine_names=True,
redefine_types=False)
self.rebuild_chains(redefine_ids=(not keep_ids), redefine_types=True, redefine_names=False)
self.rebuild_molecules(redefine_indices=False, redefine_ids=(not keep_ids), redefine_types=False,
redefine_names=True)
self.rebuild_entities(redefine_indices=True, redefine_ids=True, redefine_names=True, redefine_types=True)
del tmp_item
[docs]
def copy(self):
tmp_item = Topology()
tmp_item.atoms = self.atoms.copy()
tmp_item.groups = self.groups.copy()
tmp_item.molecules = self.molecules.copy()
tmp_item.entities = self.entities.copy()
tmp_item.components = self.components.copy()
tmp_item.chains = self.chains.copy()
tmp_item.bonds = self.bonds.copy()
return tmp_item
[docs]
def add_bonds(self, bonded_atom_pairs, skip_digestion=False):
bonded_atom_pairs = np.array(bonded_atom_pairs)
n_bonds = bonded_atom_pairs.shape[0]
aux_bonds_dataframe = Bonds_DataFrame(n_bonds=n_bonds)
aux_bonds_dataframe.atom1_index=bonded_atom_pairs[:,0]
aux_bonds_dataframe.atom2_index=bonded_atom_pairs[:,1]
df_concatenado = pd.concat([self.bonds, aux_bonds_dataframe], ignore_index=True)
self.bonds = Bonds_DataFrame(n_bonds=df_concatenado.shape[0])
self.bonds['atom1_index'] = df_concatenado['atom1_index']
self.bonds['atom2_index'] = df_concatenado['atom2_index']
self.bonds['order'] = df_concatenado['order']
self.bonds['type'] = df_concatenado['type']
self.bonds._sort_bonds()
self.bonds._remove_empty_columns()
self.rebuild_components()
del(df_concatenado, aux_bonds_dataframe)
[docs]
def remove_bonds(self, bond_indices='all', skip_digestion=False):
if is_all(bond_indices):
self.bonds = Bonds_DataFrame(n_bonds=0)
else:
self.bonds.drop(bond_indices, inplace=True)
self.bonds.reset_index(drop=True, inplace=True)
self.rebuild_components(redefine_indices=True, redefine_ids=False, redefine_names=False, redefine_types=False)
[docs]
def add_missing_bonds(self, selection='all', syntax='MolSysMT', skip_digestion=False):
from molsysmt.build import get_missing_bonds as _get_missing_bonds
bonds = _get_missing_bonds(self, selection=selection, syntax=syntax,
engine='MolSysMT', with_templates=True, with_distances=False,
skip_digestion=True)
self.add_bonds(bonds, skip_digestion=True)
self.rebuild_components(redefine_indices=True, redefine_ids=False, redefine_names=False, redefine_types=False)
[docs]
def rebuild_atoms(self, redefine_ids=True, redefine_types=True):
if redefine_ids:
self.atoms['atom_id']=np.arange(self.atoms.shape[0], dtype=int)
if redefine_types:
from molsysmt.element.atom import get_atom_type_from_atom_name
aux_dict = {}
atom_types = []
for atom_name in self.atoms['atom_name'].values:
if atom_name not in aux_dict:
atom_type=get_atom_type_from_atom_name(atom_name)
aux_dict[atom_name]=atom_type
atom_types.append(atom_type)
else:
atom_types.append(aux_dict[atom_name])
self.atoms.atom_type = np.array(atom_types, dtype=object)
del aux_dict, atom_types
[docs]
def rebuild_groups(self, redefine_ids=True, redefine_types=True):
if redefine_ids:
self.groups['group_id']=np.arange(self.groups.shape[0], dtype=int)
if redefine_types:
from molsysmt.element.group import get_group_type_from_group_name
from molsysmt.element.group.small_molecule import small_molecule_is_amino_acid
aux_dict = {}
group_types = []
for group_name in self.groups['group_name'].values:
if group_name not in aux_dict:
group_type = get_group_type_from_group_name(group_name)
if group_type == 'small molecule':
if small_molecule_is_amino_acid(self, group_name):
group_type = 'amino acid'
aux_dict[group_name]= group_type
group_types.append(group_type)
else:
group_types.append(aux_dict[group_name])
self.groups.group_type = np.array(group_types, dtype=object)
del aux_dict, group_types
[docs]
def rebuild_components(self, redefine_indices=True, redefine_ids=True, redefine_names=True, redefine_types=True):
from molsysmt.element.component import get_component_index, get_component_id, get_component_name, get_component_type
if redefine_indices:
component_index_of_atoms = get_component_index(self, element='atom', selection='all',
redefine_indices=True, skip_digestion=True)
self.atoms['component_index'] = np.array(component_index_of_atoms, dtype=int)
n_components = component_index_of_atoms[-1]+1
self.components = Components_DataFrame(n_components=n_components)
del component_index_of_atoms
if redefine_ids:
component_id_of_components = get_component_id(self, element='component', selection='all',
redefine_indices=False, redefine_ids=True,
skip_digestion=True)
self.components['component_id'] = np.array(component_id_of_components, dtype=int)
del component_id_of_components
if redefine_types:
component_type_of_components = get_component_type(self, element='component', selection='all',
redefine_indices=False, redefine_types=True,
skip_digestion=True)
self.components["component_type"] = np.array(component_type_of_components, dtype=object)
del component_type_of_components
if redefine_names:
component_name = get_component_name(self, element='component', selection='all',
redefine_indices=False,redefine_names=True,
skip_digestion=True)
self.components["component_name"] = np.array(component_name, dtype=object)
del component_name
[docs]
def rebuild_molecules(self, redefine_indices=True, redefine_ids=True, redefine_names=True, redefine_types=True,
molecules_as_components=True):
from molsysmt.element.molecule import get_molecule_index, get_molecule_id, get_molecule_name, get_molecule_type
if redefine_indices:
molecule_index_of_groups = get_molecule_index(self, element='group', selection='all',
redefine_indices=True,
skip_digestion=True)
self.groups["molecule_index"] = np.array(molecule_index_of_groups, dtype=int)
n_molecules = molecule_index_of_groups[-1]+1
self.reset_molecules(n_molecules = n_molecules)
del molecule_index_of_groups
if redefine_ids:
molecule_id_of_molecules = get_molecule_id(self, element='molecule', selection='all',
redefine_indices=False, redefine_ids=True,
skip_digestion=True)
self.molecules["molecule_id"]=np.array(molecule_id_of_molecules, dtype=int)
del molecule_id_of_molecules
if redefine_names:
molecule_name_of_molecules = get_molecule_name(self, element='molecule', selection='all',
redefine_indices=False, redefine_names=True,
skip_digestion=True)
self.molecules["molecule_name"]=np.array(molecule_name_of_molecules, dtype=object)
del molecule_name_of_molecules
if redefine_types:
molecule_type_of_molecules = get_molecule_type(self, element='molecule', selection='all',
redefine_indices=False, redefine_types=True,
skip_digestion=True)
self.molecules["molecule_type"]=np.array(molecule_type_of_molecules, dtype=object)
del molecule_type_of_molecules
[docs]
def rebuild_chains(self, redefine_indices=True, redefine_ids=True, redefine_types=True, redefine_names=True):
from molsysmt.element.chain import get_chain_index, get_chain_id, get_chain_name, get_chain_type
if redefine_indices:
chain_index_of_atoms = get_chain_index(self, element='atom', selection='all',
redefine_indices=True, skip_digestion=True)
self.atoms["chain_index"] = np.array(chain_index_of_atoms, dtype=int)
n_chains = chain_index_of_atoms[-1]+1
self.reset_chains(n_chains = n_chains)
del chain_index_of_atoms
if redefine_ids:
chain_ids_from_chain = get_chain_id(self, element='chain', selection='all',
redefine_indices=False, redefine_ids=True,
skip_digestion=True)
self.chains["chain_id"] = np.array(chain_ids_from_chain, dtype=int)
del chain_ids_from_chain
if redefine_types:
chain_types_from_chain = get_chain_type(self, element='chain', selection='all',
redefine_indices=False, redefine_types=True,
skip_digestion=True)
self.chains["chain_type"] = np.array(chain_types_from_chain, dtype=object)
del chain_types_from_chain
if redefine_names:
chain_names_from_chain = get_chain_name(self, element='chain', selection='all',
redefine_indices=False, redefine_names=True,
skip_digestion=True)
self.chains["chain_name"] = np.array(chain_names_from_chain, dtype=object)
del chain_names_from_chain
[docs]
def rebuild_entities(self, redefine_indices=True, redefine_ids=True, redefine_names=True, redefine_types=True):
from molsysmt.element.entity import get_entity_index, get_entity_id, get_entity_name, get_entity_type
if redefine_indices:
entity_index_of_molecules = get_entity_index(self, element='molecule', selection='all',
redefine_indices=True, skip_digestion=True)
self.molecules["entity_index"] = np.array(entity_index_of_molecules, dtype=int)
n_entities = entity_index_of_molecules[-1]+1
self.reset_entities(n_entities = n_entities)
del entity_index_of_molecules
if redefine_ids:
entity_ids_from_entity = get_entity_id(self, element='entity', selection='all',
redefine_indices=False, redefine_ids=True,
skip_digestion=True)
self.entities["entity_id"] = np.array(entity_ids_from_entity, dtype=int)
del entity_ids_from_entity
if redefine_names:
entity_names_from_entity = get_entity_name(self, element='entity', selection='all',
redefine_indices=False, redefine_names=True,
skip_digestion=True)
self.entities["entity_name"] = np.array(entity_names_from_entity, dtype=object)
del entity_names_from_entity
if redefine_types:
entity_types_from_entity = get_entity_type(self, element='entity', selection='all',
redefine_indices=False, redefine_types=True,
skip_digestion=True)
self.entities["entity_type"] = np.array(entity_types_from_entity, dtype=object)
del entity_types_from_entity
def _join_molecules(self, indices=None):
raise NotImplementedError
def _fix_null_values(self):
self.atoms._fix_null_values()
self.groups._fix_null_values()
self.components._fix_null_values()
self.molecules._fix_null_values()
self.entities._fix_null_values()
self.chains._fix_null_values()
self.bonds._fix_null_values()
def _sort_bonds(self):
self.bonds._sort_bonds()
[docs]
@digest()
def compare(self, item, rule='equal', output_type='boolean', skip_digestion=False, **kwargs):
if rule == 'equal':
output = {}
if 'n_atoms' in kwargs:
tmp_output = (self.atoms.shape[0]==item.atoms.shape[0])
output['n_atoms'] = (kwargs['n_atoms'] == tmp_output)
if 'atom_index' in kwargs:
tmp_output = (self.atoms.shape[0]==item.atoms.shape[0])
output['atom_index'] = (kwargs['atom_index'] == tmp_output)
if 'atom_id' in kwargs:
tmp_output = (self.atoms['atom_id'].values==item.atoms['atom_id'].values).all()
output['atom_id'] = (kwargs['atom_id'] == tmp_output)
if 'atom_name' in kwargs:
tmp_output = (self.atoms['atom_name'].values==item.atoms['atom_name'].values).all()
output['atom_name'] = (kwargs['atom_name'] == tmp_output)
if 'atom_type' in kwargs:
tmp_output = (self.atoms['atom_type'].values==item.atoms['atom_type'].values).all()
output['atom_type'] = (kwargs['atom_type'] == tmp_output)
if 'n_groups' in kwargs:
tmp_output = (self.groups.shape[0]==item.groups.shape[0])
output['n_groups'] = (kwargs['n_groups'] == tmp_output)
if 'group_index' in kwargs:
tmp_output = (self.atoms['group_index'].values==item.atoms['group_index'].values).all()
output['group_index'] = (kwargs['group_index'] == tmp_output)
if 'group_id' in kwargs:
tmp_output = (self.groups['group_id'].values==item.groups['group_id'].values).all()
output['group_id'] = (kwargs['group_id'] == tmp_output)
if 'group_name' in kwargs:
tmp_output = (self.groups['group_name'].values==item.groups['group_name'].values).all()
output['group_name'] = (kwargs['group_name'] == tmp_output)
if 'group_type' in kwargs:
tmp_output = (self.groups['group_type'].values==item.groups['group_type'].values).all()
output['group_type'] = (kwargs['group_type'] == tmp_output)
if 'component_index' in kwargs:
tmp_output = (self.atoms['component_index'].values==item.atoms['component_index'].values).all()
output['component_index'] = (kwargs['component_index'] == tmp_output)
if 'component_id' in kwargs:
tmp_output = (self.components['component_id'].values==item.components['component_id'].values).all()
output['component_id'] = (kwargs['component_id'] == tmp_output)
if 'component_name' in kwargs:
tmp_output = (self.components['component_name'].values==item.components['component_name'].values).all()
output['component_name'] = (kwargs['component_name'] == tmp_output)
if 'component_type' in kwargs:
tmp_output = (self.components['component_type'].values==item.components['component_type'].values).all()
output['component_type'] = (kwargs['component_type'] == tmp_output)
if 'molecule_index' in kwargs:
tmp_output = (self.groups['molecule_index'].values==item.groups['molecule_index'].values).all()
output['molecule_index'] = (kwargs['molecule_index'] == tmp_output)
if 'molecule_id' in kwargs:
tmp_output = (self.molecules['molecule_id'].values==item.molecules['molecule_id'].values).all()
output['molecule_id'] = (kwargs['molecule_id'] == tmp_output)
if 'molecule_name' in kwargs:
tmp_output = (self.molecules['molecule_name'].values==item.molecules['molecule_name'].values).all()
output['molecule_name'] = (kwargs['molecule_name'] == tmp_output)
if 'molecule_type' in kwargs:
tmp_output = (self.molecules['molecule_type'].values==item.molecules['molecule_type'].values).all()
output['molecule_type'] = (kwargs['molecule_type'] == tmp_output)
if 'entity_index' in kwargs:
tmp_output = (self.molecules['entity_index'].values==item.molecules['entity_index'].values).all()
output['entity_index'] = (kwargs['entity_index'] == tmp_output)
if 'entity_id' in kwargs:
tmp_output = (self.entities['entity_id'].values==item.entities['entity_id'].values).all()
output['entity_id'] = (kwargs['entity_id'] == tmp_output)
if 'entity_name' in kwargs:
tmp_output = (self.entities['entity_name'].values==item.entities['entity_name'].values).all()
output['entity_name'] = (kwargs['entity_name'] == tmp_output)
if 'entity_type' in kwargs:
tmp_output = (self.entities['entity_type'].values==item.entities['entity_type'].values).all()
output['entity_type'] = (kwargs['entity_type'] == tmp_output)
if 'chain_index' in kwargs:
tmp_output = (self.atoms['chain_index'].values==item.atoms['chain_index'].values).all()
output['chain_index'] = (kwargs['chain_index'] == tmp_output)
if 'chain_id' in kwargs:
tmp_output = (self.chains['chain_id'].values==item.chains['chain_id'].values).all()
output['chain_id'] = (kwargs['chain_id'] == tmp_output)
if 'chain_name' in kwargs:
tmp_output = (self.chains['chain_name'].values==item.chains['chain_name'].values).all()
output['chain_name'] = (kwargs['chain_name'] == tmp_output)
if 'chain_type' in kwargs:
tmp_output = (self.chains['chain_type'].values==item.chains['chain_type'].values).all()
output['chain_type'] = (kwargs['chain_type'] == tmp_output)
if 'n_bonds' in kwargs:
tmp_output = (self.bonds.shape[0]==item.bonds.shape[0]).all()
output['n_bonds'] = (kwargs['n_bonds'] == tmp_output)
if 'bonded_atom_pairs' in kwargs:
tmp_output1 = (self.bonds['atom1_index'] == item.bonds['atom1_index']).all()
tmp_output2 = (self.bonds['atom2_index'] == item.bonds['atom2_index']).all()
tmp_output = tmp_output1*tmp_output2
output['bonded_atom_pairs'] = (kwargs['bonded_atom_pairs'] == tmp_output)
if output_type=='boolean':
output = all(list(output.values()))
return output
