def read_mmcif_to_biopython(path): """Read in mmcif protein structure and report its Biopython structure Args: path (str): Path to the mmcif file. Raises: ValueError: In case _atom_site table is not present in the file. Returns: Bio.PDB.Structure.Structure: BioPython PDB structure """ if not os.path.isfile(path): raise IOError("File {} not found".format(path)) structure_builder = StructureBuilder() parsed = MMCIF2Dict().parse(path) file_name = os.path.basename(path).split(".")[0] structure_id = next((x for x in parsed), file_name).lower() structure_builder.init_structure(structure_id) try: perceived_atom_site = list(parsed.values())[0]["_atom_site"] _atom_site = _trim_models(perceived_atom_site) _parse_atom_site_biopython(_atom_site, structure_builder) except KeyError: raise ValueError("The cif file does not contain _atom_site record") return structure_builder.get_structure()
def build_structure(atoms): sb = StructureBuilder() sb.init_structure('pdb') sb.init_seg(' ') sb.init_model(0) sb.init_chain('A') i = 1 for atom in atoms: sb.init_residue('DUM', ' ', i, ' ') sb.structure[0]['A'].child_list[i - 1].add(atom) i += 1 return sb.structure
def set_structure(self, pdb_object): """Check what the user is providing and build a structure.""" # The idea here is to build missing upstream components of # the SMCRA object representation. E.g., if the user provides # a Residue, build Structure/Model/Chain. if pdb_object.level == "S": structure = pdb_object else: # Not a Structure sb = StructureBuilder() sb.init_structure("pdb") sb.init_seg(" ") if pdb_object.level == "M": sb.structure.add(pdb_object.copy()) self.structure = sb.structure else: # Not a Model sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object.copy()) else: # Not a Chain chain_id = "A" # default sb.init_chain(chain_id) if pdb_object.level == "R": # Residue # Residue extracted from a larger structure? if pdb_object.parent is not None: og_chain_id = pdb_object.parent.id sb.structure[0][chain_id].id = og_chain_id chain_id = og_chain_id sb.structure[0][chain_id].add(pdb_object.copy()) else: # Atom sb.init_residue("DUM", " ", 1, " ") # Dummy residue sb.structure[0][chain_id].child_list[0].add( pdb_object.copy()) # Fix chain identifier if Atom has grandparents. try: og_chain_id = pdb_object.parent.parent.id except AttributeError: # pdb_object.parent == None pass else: sb.structure[0][chain_id].id = og_chain_id # Return structure structure = sb.structure self.structure = structure
def pdb_writer(graph, results, dir_name="templates"): PDB_CODE = graph.protein_name structure = graph.structure path = "%s/%s/" % (dir_name, PDB_CODE) if len(results) > 0: if not os.path.exists(path): os.makedirs(path) else: print 'No results' return top = min(NUM_OF_RESULTS_TO_SAVE, len(results)) i = 0 for key, value in results[:top]: # TODO - save as different chains sb = StructureBuilder() sb.init_structure(PDB_CODE) sb.init_model(0) sb.init_chain('X') for node_id in key[0]: for res in graph.nodes[node_id].residues: sb.init_seg(res.get_segid()) sb.init_residue(res.get_resname(), res.get_id()[0], res.get_id()[1], res.get_id()[2]) for atom in res: sb.init_atom(atom.get_name(), atom.get_coord(), atom.get_bfactor(), atom.get_occupancy(), atom.get_altloc(), atom.get_fullname()) sb.init_chain('Y') for node_id in key[1]: for res in graph.nodes[node_id].residues: sb.init_residue(res.get_resname(), res.get_id()[0], res.get_id()[1], res.get_id()[2]) for atom in res: sb.init_atom(atom.get_name(), atom.get_coord(), atom.get_bfactor(), atom.get_occupancy(), atom.get_altloc(), atom.get_fullname()) filename = path + "interface%d.pdb" % (i) io = PDBIO() io.set_structure(sb.get_structure()) io.save(filename) i += 1 return 1
def residues_to_struct(name_residues_list, structID): """ Build a structure from a list of (chain name, residues) - each as a chain. """ builder = StructureBuilder() builder.init_structure(structID) builder.init_model(0) for (name, residues) in name_residues_list: builder.init_chain(name) for res in residues: builder.chain.add(res) return builder.get_structure()
def coarse_grain(self, cg_type="CA_TRACE"): """ Reduces the protein structure complexity to a few (pseudo-)atoms per residue. Parameters: - cg_type: CA_TRACE (Ca-only) [Default] ENCAD_3P (CA, O, SC Beads) MARTINI (CA, O, SC Beads) Returns a new structure object. """ # Import CG Types import CG_Models CG_Library = { "CA_TRACE": CG_Models.CA_TRACE, "ENCAD_3P": CG_Models.ENCAD_3P, "MARTINI": CG_Models.MARTINI } CG_Method = CG_Library[cg_type] # Creates a brand new structure object from Bio.PDB.StructureBuilder import StructureBuilder structure_builder = StructureBuilder() cg_id = "CG_" + self.id structure_builder.init_structure(cg_id) structure_builder.init_seg(' ') # Empty SEGID for model in self: structure_builder.init_model(model.id) for chain in model: structure_builder.init_chain(chain.id) cur_chain = structure_builder.chain for residue in chain: cg_residue = CG_Method(residue) cur_chain.add(cg_residue) cg_structure = structure_builder.get_structure() return cg_structure
class XBGFParser: def __init__(self, PERMISSIVE=1, structure_builder=None): if structure_builder != None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.PERMISSIVE = PERMISSIVE # public interface def parse(self, id, file): self.structure_builder.init_structure(id) if isinstance(file, basestring): file=open(file) self.charges = dict() self.chain_suffix = 0 self._parse(file.readlines()) self.structure = self.structure_builder.get_structure() return self._process_structure() # private methods def _parse(self, lines): self.structure_builder.init_model(0) self.structure_builder.init_seg("") self.current_chain_id = None self.current_residue_id = None self.current_resname = None for i in range(0, len(lines)): self.line_counter = i + 1 self.structure_builder.set_line_counter(self.line_counter) line = lines[i] if line[0:6] == 'ATOM ': self._update_atom(line) def _update_chain(self, line): chain_id = self._extract_chain(line) if self.current_chain_id != chain_id: try: self.structure_builder.init_chain(chain_id) self.current_chain_id = chain_id self.current_residue_id = None self.current_resname = None except PDBConstructionException, message: self._handle_PDB_exception(message)
class XBGFParser: def __init__(self, PERMISSIVE=1, structure_builder=None): if structure_builder != None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.PERMISSIVE = PERMISSIVE # public interface def parse(self, id, file): self.structure_builder.init_structure(id) if isinstance(file, basestring): file = open(file) self.charges = dict() self.chain_suffix = 0 self._parse(file.readlines()) self.structure = self.structure_builder.get_structure() return self._process_structure() # private methods def _parse(self, lines): self.structure_builder.init_model(0) self.structure_builder.init_seg("") self.current_chain_id = None self.current_residue_id = None self.current_resname = None for i in range(0, len(lines)): self.line_counter = i + 1 self.structure_builder.set_line_counter(self.line_counter) line = lines[i] if line[0:6] == 'ATOM ': self._update_atom(line) def _update_chain(self, line): chain_id = self._extract_chain(line) if self.current_chain_id != chain_id: try: self.structure_builder.init_chain(chain_id) self.current_chain_id = chain_id self.current_residue_id = None self.current_resname = None except PDBConstructionException, message: self._handle_PDB_exception(message)
def write_helical_axes(self, filename): """Writes helical axes in PDB format.""" sb = StructureBuilder() sb.init_structure('') sb.init_model('') for cpos, chain in enumerate(self.chains): sb.init_chain(str(cpos)) sb.init_seg('') for pos, i in enumerate(chain.res[1:-1]): sb.init_residue('ALA', ' ', pos, ' ') sb.init_atom('CA', i.O._ar, 0, 0, " ", ' CA ', 1) io = PDBIO() io.set_structure(sb.structure) io.save(filename)
def set_structure(self, pdb_object): """Check what object the user is providing and build a structure.""" # This is duplicated from the PDBIO class if pdb_object.level == "S": structure = pdb_object else: sb = StructureBuilder() sb.init_structure('pdb') sb.init_seg(' ') # Build parts as necessary if pdb_object.level == "M": sb.structure.add(pdb_object) self.structure = sb.structure else: sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object) else: sb.init_chain('A') if pdb_object.level == "R": try: parent_id = pdb_object.parent.id sb.structure[0]['A'].id = parent_id except ValueError: pass sb.structure[0]['A'].add(pdb_object) else: # Atom sb.init_residue('DUM', ' ', 1, ' ') try: parent_id = pdb_object.parent.parent.id sb.structure[0]['A'].id = parent_id except ValueError: pass sb.structure[0]['A'].child_list[0].add(pdb_object) # Return structure structure = sb.structure self.structure = structure
def set_structure(self, pdb_object): """Check what the user is providing and build a structure.""" if pdb_object.level == "S": structure = pdb_object else: sb = StructureBuilder() sb.init_structure("pdb") sb.init_seg(" ") # Build parts as necessary if pdb_object.level == "M": sb.structure.add(pdb_object.copy()) self.structure = sb.structure else: sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object.copy()) else: sb.init_chain("A") if pdb_object.level == "R": try: parent_id = pdb_object.parent.id sb.structure[0]["A"].id = parent_id except Exception: pass sb.structure[0]["A"].add(pdb_object.copy()) else: # Atom sb.init_residue("DUM", " ", 1, " ") try: parent_id = pdb_object.parent.parent.id sb.structure[0]["A"].id = parent_id except Exception: pass sb.structure[0]["A"].child_list[0].add( pdb_object.copy()) # Return structure structure = sb.structure self.structure = structure
def set_structure(self, pdb_object): """Check what the user is providing and build a structure.""" if pdb_object.level == "S": structure = pdb_object else: sb = StructureBuilder() sb.init_structure('pdb') sb.init_seg(' ') # Build parts as necessary if pdb_object.level == "M": sb.structure.add(pdb_object.copy()) self.structure = sb.structure else: sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object.copy()) else: sb.init_chain('A') if pdb_object.level == "R": try: parent_id = pdb_object.parent.id sb.structure[0]['A'].id = parent_id except Exception: pass sb.structure[0]['A'].add(pdb_object.copy()) else: # Atom sb.init_residue('DUM', ' ', 1, ' ') try: parent_id = pdb_object.parent.parent.id sb.structure[0]['A'].id = parent_id except Exception: pass sb.structure[0]['A'].child_list[0].add(pdb_object.copy()) # Return structure structure = sb.structure self.structure = structure
def set_structure(self, pdb_object): # Check what the user is providing and build a structure appropriately if pdb_object.level == "S": structure = pdb_object else: sb = StructureBuilder() sb.init_structure('pdb') sb.init_seg(' ') # Build parts as necessary if pdb_object.level == "M": sb.structure.add(pdb_object) self.structure = sb.structure else: sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object) else: sb.init_chain('A') if pdb_object.level == "R": try: parent_id = pdb_object.parent.id sb.structure[0]['A'].id = parent_id except Exception: pass sb.structure[0]['A'].add(pdb_object) else: # Atom sb.init_residue('DUM', ' ', 1, ' ') try: parent_id = pdb_object.parent.parent.id sb.structure[0]['A'].id = parent_id except Exception: pass sb.structure[0]['A'].child_list[0].add(pdb_object) # Return structure structure = sb.structure self.structure = structure
def set_structure(self, pdb_object): # Check what the user is providing and build a structure appropriately if pdb_object.level == "S": structure = pdb_object else: sb = StructureBuilder() sb.init_structure("pdb") sb.init_seg(" ") # Build parts as necessary if pdb_object.level == "M": sb.structure.add(pdb_object) self.structure = sb.structure else: sb.init_model(0) if pdb_object.level == "C": sb.structure[0].add(pdb_object) else: sb.init_chain("A") if pdb_object.level == "R": try: parent_id = pdb_object.parent.id sb.structure[0]["A"].id = parent_id except Exception: pass sb.structure[0]["A"].add(pdb_object) else: # Atom sb.init_residue("DUM", " ", 1, " ") try: parent_id = pdb_object.parent.parent.id sb.structure[0]["A"].id = parent_id except Exception: pass sb.structure[0]["A"].child_list[0].add(pdb_object) # Return structure structure = sb.structure self.structure = structure
class StructureDecoder(object): """Class to pass the data from mmtf-python into a Biopython data structure.""" def __init__(self): """Initialize the class.""" self.this_type = "" def init_structure(self, total_num_bonds, total_num_atoms, total_num_groups, total_num_chains, total_num_models, structure_id): """Initialize the structure object. :param total_num_bonds: the number of bonds in the structure :param total_num_atoms: the number of atoms in the structure :param total_num_groups: the number of groups in the structure :param total_num_chains: the number of chains in the structure :param total_num_models: the number of models in the structure :param structure_id: the id of the structure (e.g. PDB id) """ self.structure_bulder = StructureBuilder() self.structure_bulder.init_structure(structure_id=structure_id) self.chain_index_to_type_map = {} self.chain_index_to_seq_map = {} self.chain_index_to_description_map = {} self.chain_counter = 0 def set_atom_info(self, atom_name, serial_number, alternative_location_id, x, y, z, occupancy, temperature_factor, element, charge): """Create an atom object an set the information. :param atom_name: the atom name, e.g. CA for this atom :param serial_number: the serial id of the atom (e.g. 1) :param alternative_location_id: the alternative location id for the atom, if present :param x: the x coordiante of the atom :param y: the y coordinate of the atom :param z: the z coordinate of the atom :param occupancy: the occupancy of the atom :param temperature_factor: the temperature factor of the atom :param element: the element of the atom, e.g. C for carbon. According to IUPAC. Calcium is Ca :param charge: the formal atomic charge of the atom """ # MMTF uses "\x00" (the NUL character) to indicate to altloc, so convert # that to the space required by StructureBuilder if alternative_location_id == "\x00": alternative_location_id = " " # Atom_name is in twice - the full_name is with spaces self.structure_bulder.init_atom(str(atom_name), [x, y, z], temperature_factor, occupancy, alternative_location_id, str(atom_name), serial_number=serial_number, element=str(element).upper()) def set_chain_info(self, chain_id, chain_name, num_groups): """Set the chain information. :param chain_id: the asym chain id from mmCIF :param chain_name: the auth chain id from mmCIF :param num_groups: the number of groups this chain has """ # A Bradley - chose to use chain_name (auth_id) as it complies # with current Biopython. Chain_id might be better. self.structure_bulder.init_chain(chain_id=chain_name) if self.chain_index_to_type_map[self.chain_counter] == "polymer": self.this_type = " " elif self.chain_index_to_type_map[self.chain_counter] == "non-polymer": self.this_type = "H" elif self.chain_index_to_type_map[self.chain_counter] == "water": self.this_type = "W" self.chain_counter += 1 def set_entity_info(self, chain_indices, sequence, description, entity_type): """Set the entity level information for the structure. :param chain_indices: the indices of the chains for this entity :param sequence: the one letter code sequence for this entity :param description: the description for this entity :param entity_type: the entity type (polymer,non-polymer,water) """ for chain_ind in chain_indices: self.chain_index_to_type_map[chain_ind] = entity_type self.chain_index_to_seq_map[chain_ind] = sequence self.chain_index_to_description_map[chain_ind] = description def set_group_info(self, group_name, group_number, insertion_code, group_type, atom_count, bond_count, single_letter_code, sequence_index, secondary_structure_type): """Set the information for a group :param group_name: the name of this group, e.g. LYS :param group_number: the residue number of this group :param insertion_code: the insertion code for this group :param group_type: a string indicating the type of group (as found in the chemcomp dictionary. Empty string if none available. :param atom_count: the number of atoms in the group :param bond_count: the number of unique bonds in the group :param single_letter_code: the single letter code of the group :param sequence_index: the index of this group in the sequence defined by the entity :param secondary_structure_type: the type of secondary structure used (types are according to DSSP and number to type mappings are defined in the specification) """ # MMTF uses a NUL character to indicate a blank insertion code, but # StructureBuilder expects a space instead. if insertion_code == "\x00": insertion_code = " " self.structure_bulder.init_seg(' ') self.structure_bulder.init_residue(group_name, self.this_type, group_number, insertion_code) def set_model_info(self, model_id, chain_count): """Set the information for a model. :param model_id: the index for the model :param chain_count: the number of chains in the model """ self.structure_bulder.init_model(model_id) def set_xtal_info(self, space_group, unit_cell): """Set the crystallographic information for the structure. :param space_group: the space group name, e.g. "P 21 21 21" :param unit_cell: an array of length 6 with the unit cell parameters in order: a, b, c, alpha, beta, gamma """ self.structure_bulder.set_symmetry(space_group, unit_cell) def set_header_info(self, r_free, r_work, resolution, title, deposition_date, release_date, experimnetal_methods): """Sets the header information. :param r_free: the measured R-Free for the structure :param r_work: the measure R-Work for the structure :param resolution: the resolution of the structure :param title: the title of the structure :param deposition_date: the deposition date of the structure :param release_date: the release date of the structure :param experimnetal_methods: the list of experimental methods in the structure """ pass def set_bio_assembly_trans(self, bio_assembly_index, input_chain_indices, input_transform): """Set the Bioassembly transformation information. A single bioassembly can have multiple transforms. :param bio_assembly_index: the integer index of the bioassembly :param input_chain_indices: the list of integer indices for the chains of this bioassembly :param input_transform: the list of doubles for the transform of this bioassmbly transform. """ pass def finalize_structure(self): """Any functions needed to cleanup the structure.""" pass def set_group_bond(self, atom_index_one, atom_index_two, bond_order): """Add bonds within a group. :param atom_index_one: the integer atom index (in the group) of the first partner in the bond :param atom_index_two: the integer atom index (in the group) of the second partner in the bond :param bond_order: the integer bond order """ pass def set_inter_group_bond(self, atom_index_one, atom_index_two, bond_order): """Add bonds between groups. :param atom_index_one: the integer atom index (in the structure) of the first partner in the bond :param atom_index_two: the integer atom index (in the structure) of the second partner in the bond :param bond_order: the bond order """ pass
class StructureDecoder(object): """Class to pass the data from mmtf-python into a Biopython data structure.""" def __init__(self): """Initialize the class.""" self.this_type = "" def init_structure(self, total_num_bonds, total_num_atoms, total_num_groups, total_num_chains, total_num_models, structure_id): """Initialize the structure object. :param total_num_bonds: the number of bonds in the structure :param total_num_atoms: the number of atoms in the structure :param total_num_groups: the number of groups in the structure :param total_num_chains: the number of chains in the structure :param total_num_models: the number of models in the structure :param structure_id: the id of the structure (e.g. PDB id) """ self.structure_bulder = StructureBuilder() self.structure_bulder.init_structure(structure_id=structure_id) self.chain_index_to_type_map = {} self.chain_index_to_seq_map = {} self.chain_index_to_description_map = {} self.chain_counter = 0 def set_atom_info(self, atom_name, serial_number, alternative_location_id, x, y, z, occupancy, temperature_factor, element, charge): """Create an atom object an set the information. :param atom_name: the atom name, e.g. CA for this atom :param serial_number: the serial id of the atom (e.g. 1) :param alternative_location_id: the alternative location id for the atom, if present :param x: the x coordiante of the atom :param y: the y coordinate of the atom :param z: the z coordinate of the atom :param occupancy: the occupancy of the atom :param temperature_factor: the temperature factor of the atom :param element: the element of the atom, e.g. C for carbon. According to IUPAC. Calcium is Ca :param charge: the formal atomic charge of the atom """ # MMTF uses "\x00" (the NUL character) to indicate to altloc, so convert # that to the space required by StructureBuilder if alternative_location_id == "\x00": alternative_location_id = " " # Atom_name is in twice - the full_name is with spaces self.structure_bulder.init_atom(str(atom_name), numpy.array((x, y, z), "f"), temperature_factor, occupancy, alternative_location_id, str(atom_name), serial_number=serial_number, element=str(element).upper()) def set_chain_info(self, chain_id, chain_name, num_groups): """Set the chain information. :param chain_id: the asym chain id from mmCIF :param chain_name: the auth chain id from mmCIF :param num_groups: the number of groups this chain has """ # A Bradley - chose to use chain_name (auth_id) as it complies # with current Biopython. Chain_id might be better. self.structure_bulder.init_chain(chain_id=chain_name) if self.chain_index_to_type_map[self.chain_counter] == "polymer": self.this_type = " " elif self.chain_index_to_type_map[self.chain_counter] == "non-polymer": self.this_type = "H" elif self.chain_index_to_type_map[self.chain_counter] == "water": self.this_type = "W" self.chain_counter += 1 def set_entity_info(self, chain_indices, sequence, description, entity_type): """Set the entity level information for the structure. :param chain_indices: the indices of the chains for this entity :param sequence: the one letter code sequence for this entity :param description: the description for this entity :param entity_type: the entity type (polymer,non-polymer,water) """ for chain_ind in chain_indices: self.chain_index_to_type_map[chain_ind] = entity_type self.chain_index_to_seq_map[chain_ind] = sequence self.chain_index_to_description_map[chain_ind] = description def set_group_info(self, group_name, group_number, insertion_code, group_type, atom_count, bond_count, single_letter_code, sequence_index, secondary_structure_type): """Set the information for a group. :param group_name: the name of this group, e.g. LYS :param group_number: the residue number of this group :param insertion_code: the insertion code for this group :param group_type: a string indicating the type of group (as found in the chemcomp dictionary. Empty string if none available. :param atom_count: the number of atoms in the group :param bond_count: the number of unique bonds in the group :param single_letter_code: the single letter code of the group :param sequence_index: the index of this group in the sequence defined by the entity :param secondary_structure_type: the type of secondary structure used (types are according to DSSP and number to type mappings are defined in the specification) """ # MMTF uses a NUL character to indicate a blank insertion code, but # StructureBuilder expects a space instead. if insertion_code == "\x00": insertion_code = " " self.structure_bulder.init_seg(' ') self.structure_bulder.init_residue(group_name, self.this_type, group_number, insertion_code) def set_model_info(self, model_id, chain_count): """Set the information for a model. :param model_id: the index for the model :param chain_count: the number of chains in the model """ self.structure_bulder.init_model(model_id) def set_xtal_info(self, space_group, unit_cell): """Set the crystallographic information for the structure. :param space_group: the space group name, e.g. "P 21 21 21" :param unit_cell: an array of length 6 with the unit cell parameters in order: a, b, c, alpha, beta, gamma """ self.structure_bulder.set_symmetry(space_group, unit_cell) def set_header_info(self, r_free, r_work, resolution, title, deposition_date, release_date, experimnetal_methods): """Set the header information. :param r_free: the measured R-Free for the structure :param r_work: the measure R-Work for the structure :param resolution: the resolution of the structure :param title: the title of the structure :param deposition_date: the deposition date of the structure :param release_date: the release date of the structure :param experimnetal_methods: the list of experimental methods in the structure """ pass def set_bio_assembly_trans(self, bio_assembly_index, input_chain_indices, input_transform): """Set the Bioassembly transformation information. A single bioassembly can have multiple transforms. :param bio_assembly_index: the integer index of the bioassembly :param input_chain_indices: the list of integer indices for the chains of this bioassembly :param input_transform: the list of doubles for the transform of this bioassmbly transform. """ pass def finalize_structure(self): """Any functions needed to cleanup the structure.""" pass def set_group_bond(self, atom_index_one, atom_index_two, bond_order): """Add bonds within a group. :param atom_index_one: the integer atom index (in the group) of the first partner in the bond :param atom_index_two: the integer atom index (in the group) of the second partner in the bond :param bond_order: the integer bond order """ pass def set_inter_group_bond(self, atom_index_one, atom_index_two, bond_order): """Add bonds between groups. :param atom_index_one: the integer atom index (in the structure) of the first partner in the bond :param atom_index_two: the integer atom index (in the structure) of the second partner in the bond :param bond_order: the bond order """ pass
class PDBP_read(object): def __init__(self, get_header=False, structure_builder=None, PERMISSIVE=True): """arguments: PERMISSIVE, Evaluated as a Boolean. If ture, the exception are caught, some residues or atoms will be missing.THESE EXCEPTIONS ARE DUE TO PROBLEMS IN THE PDB FILE! structure_builder, an optional user implemented StructureBuilder class. """ #get a structure_builder class if structure_builder is not None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.trailer = None self.line_counter = 0 self.PERMISSIVE = bool(PERMISSIVE) def get_structure(self, id, file): """return the structure. argurements: -id - the name of the sturecture -file - pdb filename """ self.header = None self.trailer = None #make a StructureBuilder instance self.structure_builder.init_structure(id) if file[-2:] == 'gz': #try: #with open(file,'r+',encoding='utf-8') as handle: #self._parse(handle.readlines()) with gzip.open(file, 'r') as handle: #######按照行读取pdb文件 self._parse(handle.read().decode("utf-8").split('\n')) else: try: with open(file, 'r') as handle: self._parse(handle.readlines()) except Exception: print("%s cannot be open!" % (file)) #exit() #??????? self.structure_builder.set_header(self.header) # return the structure instance structure = self.structure_builder.get_structure() return structure def get_header(self): """return the header""" return self.header def get_trailer(self): """return the trailer""" return self.trailer #private methods def _parse(self, header_coords_trailer): """parser the pdb file(private)""" self.coords_trailer = self._get_header(header_coords_trailer) ## parse the atomicdata; return the pdb file triler self.trailer = self._parse_coordinates(self.coords_trailer) def _get_header(self, header_coords_trailer): """get the header of the pdb file""" structure_builder = self.structure_builder i = 0 line_nums = len(header_coords_trailer) ''' for index, line in enumerate(header_coords_trailer): print(index, line) print(line_nums, type(header_coords_trailer)) ''' for i in range(0, line_nums): structure_builder.set_line_counter(i + 1) line = header_coords_trailer[i] record_type = line[0:6] if record_type == "ATOM " or record_type == "HETATM" or record_type == "MODEL ": break #header = header_coords_trailer[0:i] #return the rest of the coodstrailer self.line_counter = i coords_trailer = header_coords_trailer[i:] #header_dict = self._parse_pdb_header_list(header) return coords_trailer def _parse_coordinates(self, coords_trailer): """parse the atomic data in teh PDB file """ local_line_counter = 0 #n=0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None lines_num1 = len(coords_trailer) for i in range(0, lines_num1): line = coords_trailer[i].rstrip('\n') record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 # the all lines nums include header coods and trailer structure_builder.set_line_counter(global_line_counter) if record_type == "ATOM ": #or record_type == "HETATM": #Initialize the Model - there was no explicit MODEL record if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # a atom has several species, eg "N B" name = fullname else: #eg: "CA" name = split_list[0] altloc = line[16] resname = line[17:20] chainid = line[21] try: serial_number = int(line[6:11]) except Exception: serial_number = 0 resseq = int(line[22:26].split()[0]) icode = line[26] if record_type == "HETATM": if resname == "HOH" or resname == "WAT": hetero_flag = "W" else: hetero_flag = "H" else: hetero_flag = " " residue_id = (hetero_flag, resseq, icode) try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except Exception: raise PDBConstructionException( "Invalid or missing coordinate(s) at line %i." % global_line_counter) coord = numpy.array((x, y, z), "f") try: occupancy = float(line[54:60]) except Exception: self._handle_PDB_exception("Invalid or missing occupancy", global_line_counter) #occupancy = None # Rather than arbitrary zero or one occupancy = None # Rather than arbitrary zero or one if occupancy is not None and occupancy < 0: warnings.warn("Negative occupancy in one or more atoms", PDBConstructionWarning) try: bfactor = float(line[60:66]) except Exception: self._handle_PDB_exception("Invalid or missing B factor", global_line_counter) bfactor = 0.0 # The PDB use a default of zero if the data is missing segid = line[72:76] element = line[76:78].strip().upper() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif current_residue_id != residue_id or current_resname != resname: current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) # init atom try: structure_builder.init_atom(name, coord, bfactor, occupancy, altloc, fullname, serial_number, element) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif record_type == "MODEL ": try: serial_num = int(line[10:14]) except Exception: self._handle_PDB_exception( "Invalid or missing model serial number", global_line_counter) serial_num = 0 structure_builder.init_model(current_model_id, serial_num) current_model_id += 1 model_open = 1 current_chain_id = None current_residue_id = None elif record_type == "END " or record_type == "CONECT": # End of atomic data, return the trailer self.line_counter += local_line_counter return coords_trailer[local_line_counter:] elif record_type == "ENDMDL": model_open = 0 current_chain_id = None current_residue_id = None elif record_type == "SIGUIJ": # standard deviation of anisotropic B factor siguij = [ float(x) for x in (line[28:35], line[35:42], line[42:49], line[49:56], line[56:63], line[63:70]) ] # U sigma's are scaled by 10^4 siguij_array = (numpy.array(siguij, "f") / 10000.0).astype("f") structure_builder.set_siguij(siguij_array) elif record_type == "SIGATM": # standard deviation of atomic positions sigatm = [ float(x) for x in (line[30:38], line[38:45], line[46:54], line[54:60], line[60:66]) ] sigatm_array = numpy.array(sigatm, "f") structure_builder.set_sigatm(sigatm_array) local_line_counter += 1 # EOF (does not end in END or CONECT) self.line_counter = self.line_counter + local_line_counter return [] #info = (resname, resseq, serial_number,fullname, coord) #yield info def _handle_PDB_exception(self, message, line_counter): message = "%s at line %i." % (message, line_counter) if self.PERMISSIVE: # just print a warning - some residues/atoms may be missing warnings.warn( "PDBConstructionException: %s\n" "Exception ignored.\n" "Some atoms or residues may be missing in the data structure." % message, PDBConstructionWarning) else: # exceptions are fatal - raise again with new message (including line nr) raise PDBConstructionException(message)
class PDBParser(object): """ Parse a PDB file and return a Structure object. """ def __init__(self, PERMISSIVE=True, get_header=False, structure_builder=None, QUIET=False): """ The PDB parser call a number of standard methods in an aggregated StructureBuilder object. Normally this object is instanciated by the PDBParser object itself, but if the user provides his own StructureBuilder object, the latter is used instead. Arguments: o PERMISSIVE - Evaluated as a Boolean. If false, exceptions in constructing the SMCRA data structure are fatal. If true (DEFAULT), the exceptions are caught, but some residues or atoms will be missing. THESE EXCEPTIONS ARE DUE TO PROBLEMS IN THE PDB FILE!. o structure_builder - an optional user implemented StructureBuilder class. o QUIET - Evaluated as a Boolean. If true, warnings issued in constructing the SMCRA data will be supressed. If false (DEFAULT), they will be shown. These warnings might be indicative of problems in the PDB file! """ if structure_builder != None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.trailer = None self.line_counter = 0 self.PERMISSIVE = bool(PERMISSIVE) self.QUIET = bool(QUIET) # Public methods def get_structure(self, id, file): """Return the structure. Arguments: o id - string, the id that will be used for the structure o file - name of the PDB file OR an open filehandle """ if self.QUIET: warning_list = warnings.filters[:] warnings.filterwarnings('ignore', category=PDBConstructionWarning) self.header = None self.trailer = None # Make a StructureBuilder instance (pass id of structure as parameter) self.structure_builder.init_structure(id) with as_handle(file) as handle: self._parse(handle.readlines()) self.structure_builder.set_header(self.header) # Return the Structure instance structure = self.structure_builder.get_structure() if self.QUIET: warnings.filters = warning_list return structure def get_header(self): "Return the header." return self.header def get_trailer(self): "Return the trailer." return self.trailer # Private methods def _parse(self, header_coords_trailer): "Parse the PDB file." # Extract the header; return the rest of the file self.header, coords_trailer = self._get_header(header_coords_trailer) # Parse the atomic data; return the PDB file trailer self.trailer = self._parse_coordinates(coords_trailer) def _get_header(self, header_coords_trailer): "Get the header of the PDB file, return the rest." structure_builder = self.structure_builder i = 0 for i in range(0, len(header_coords_trailer)): structure_builder.set_line_counter(i + 1) line = header_coords_trailer[i] record_type = line[0:6] if (record_type == 'ATOM ' or record_type == 'HETATM' or record_type == 'MODEL '): break header = header_coords_trailer[0:i] # Return the rest of the coords+trailer for further processing self.line_counter = i coords_trailer = header_coords_trailer[i:] header_dict = _parse_pdb_header_list(header) return header_dict, coords_trailer def _parse_coordinates(self, coords_trailer): "Parse the atomic data in the PDB file." local_line_counter = 0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None for i in range(0, len(coords_trailer)): line = coords_trailer[i] record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 structure_builder.set_line_counter(global_line_counter) if (record_type == 'ATOM ' or record_type == 'HETATM'): # Initialize the Model - there was no explicit MODEL record if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # atom name has internal spaces, e.g. " N B ", so # we do not strip spaces name = fullname else: # atom name is like " CA ", so we can strip spaces name = split_list[0] altloc = line[16:17] resname = line[17:20] chainid = line[21:22] try: serial_number = int(line[6:11]) except: serial_number = 0 resseq = int(line[22:26].split()[0]) # sequence identifier icode = line[26:27] # insertion code if record_type == 'HETATM': # hetero atom flag if resname == "HOH" or resname == "WAT": hetero_flag = "W" else: hetero_flag = "H" else: hetero_flag = " " residue_id = (hetero_flag, resseq, icode) # atomic coordinates try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except: #Should we allow parsing to continue in permissive mode? #If so what coordindates should we default to? Easier to abort! raise PDBConstructionException(\ "Invalid or missing coordinate(s) at line %i." \ % global_line_counter) coord = numpy.array((x, y, z), 'f') # occupancy & B factor try: occupancy = float(line[54:60]) except: self._handle_PDB_exception("Invalid or missing occupancy", global_line_counter) occupancy = 0.0 #Is one or zero a good default? try: bfactor = float(line[60:66]) except: self._handle_PDB_exception("Invalid or missing B factor", global_line_counter) bfactor = 0.0 #The PDB use a default of zero if the data is missing segid = line[72:76] element = line[76:78].strip() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter) elif current_residue_id != residue_id or current_resname != resname: current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter) # init atom try: structure_builder.init_atom(name, coord, bfactor, occupancy, altloc, fullname, serial_number, element) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter)
class PDBParser(object): """ Parse a PDB file and return a Structure object. """ def __init__(self, PERMISSIVE=True, get_header=False, structure_builder=None, QUIET=False): """ The PDB parser call a number of standard methods in an aggregated StructureBuilder object. Normally this object is instanciated by the PDBParser object itself, but if the user provides his/her own StructureBuilder object, the latter is used instead. Arguments: o PERMISSIVE - Evaluated as a Boolean. If false, exceptions in constructing the SMCRA data structure are fatal. If true (DEFAULT), the exceptions are caught, but some residues or atoms will be missing. THESE EXCEPTIONS ARE DUE TO PROBLEMS IN THE PDB FILE!. o structure_builder - an optional user implemented StructureBuilder class. o QUIET - Evaluated as a Boolean. If true, warnings issued in constructing the SMCRA data will be suppressed. If false (DEFAULT), they will be shown. These warnings might be indicative of problems in the PDB file! """ if structure_builder is not None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.trailer = None self.line_counter = 0 self.PERMISSIVE = bool(PERMISSIVE) self.QUIET = bool(QUIET) # Public methods def get_structure(self, id, file): """Return the structure. Arguments: o id - string, the id that will be used for the structure o file - name of the PDB file OR an open filehandle """ if self.QUIET: warning_list = warnings.filters[:] warnings.filterwarnings("ignore", category=PDBConstructionWarning) self.header = None self.trailer = None # Make a StructureBuilder instance (pass id of structure as parameter) self.structure_builder.init_structure(id) with as_handle(file) as handle: self._parse(handle.readlines()) self.structure_builder.set_header(self.header) # Return the Structure instance structure = self.structure_builder.get_structure() if self.QUIET: warnings.filters = warning_list return structure def get_header(self): "Return the header." return self.header def get_trailer(self): "Return the trailer." return self.trailer # Private methods def _parse(self, header_coords_trailer): "Parse the PDB file." # Extract the header; return the rest of the file self.header, coords_trailer = self._get_header(header_coords_trailer) # Parse the atomic data; return the PDB file trailer self.trailer = self._parse_coordinates(coords_trailer) def _get_header(self, header_coords_trailer): "Get the header of the PDB file, return the rest." structure_builder = self.structure_builder i = 0 for i in range(0, len(header_coords_trailer)): structure_builder.set_line_counter(i + 1) line = header_coords_trailer[i] record_type = line[0:6] if record_type == "ATOM " or record_type == "HETATM" or record_type == "MODEL ": break header = header_coords_trailer[0:i] # Return the rest of the coords+trailer for further processing self.line_counter = i coords_trailer = header_coords_trailer[i:] header_dict = _parse_pdb_header_list(header) return header_dict, coords_trailer def _parse_coordinates(self, coords_trailer): "Parse the atomic data in the PDB file." local_line_counter = 0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None for i in range(0, len(coords_trailer)): line = coords_trailer[i] record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 structure_builder.set_line_counter(global_line_counter) if record_type == "ATOM " or record_type == "HETATM": # Initialize the Model - there was no explicit MODEL record if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # atom name has internal spaces, e.g. " N B ", so # we do not strip spaces name = fullname else: # atom name is like " CA ", so we can strip spaces name = split_list[0] altloc = line[16] resname = line[17:20] chainid = line[21] try: serial_number = int(line[6:11]) except: serial_number = 0 resseq = int(line[22:26].split()[0]) # sequence identifier icode = line[26] # insertion code if record_type == "HETATM": # hetero atom flag if resname == "HOH" or resname == "WAT": hetero_flag = "W" else: hetero_flag = "H" else: hetero_flag = " " residue_id = (hetero_flag, resseq, icode) # atomic coordinates try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except: # Should we allow parsing to continue in permissive mode? # If so, what coordinates should we default to? Easier to abort! raise PDBConstructionException("Invalid or missing coordinate(s) at line %i." % global_line_counter) coord = numpy.array((x, y, z), "f") # occupancy & B factor try: occupancy = float(line[54:60]) except: self._handle_PDB_exception("Invalid or missing occupancy", global_line_counter) occupancy = None # Rather than arbitrary zero or one try: bfactor = float(line[60:66]) except: self._handle_PDB_exception("Invalid or missing B factor", global_line_counter) bfactor = 0.0 # The PDB use a default of zero if the data is missing segid = line[72:76] element = line[76:78].strip() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter) elif current_residue_id != residue_id or current_resname != resname: current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter) # init atom try: structure_builder.init_atom(name, coord, bfactor, occupancy, altloc, fullname, serial_number, element) except PDBConstructionException, message: self._handle_PDB_exception(message, global_line_counter)
'F': 9 # Fibril } # Load things P = PDBParser() # P = PDBParser() io = PDBIO() # Parse PDB and run DSSP pdbf_path = os.path.abspath(sys.argv[1]) aa_model = P.get_structure('aa_model', pdbf_path) # Convert to MARTINI types # Assign by chain and build the cg structure already structure_builder = StructureBuilder() structure_builder.init_structure("cg_model") structure_builder.init_seg(' ') # Empty SEGID nbeads = 0 for model in aa_model: structure_builder.init_model(model.id) dssp = DSSP(model, pdbf_path) for chain in model: structure_builder.init_chain(chain.id) # Get SS information and translate it to MARTINI dssp_ss = [] for residue in chain:
class PDBParser(object): """Parse a PDB file and return a Structure object.""" def __init__(self, PERMISSIVE=True, get_header=False, structure_builder=None, QUIET=False): """Create a PDBParser object. The PDB parser call a number of standard methods in an aggregated StructureBuilder object. Normally this object is instanciated by the PDBParser object itself, but if the user provides his/her own StructureBuilder object, the latter is used instead. Arguments: - PERMISSIVE - Evaluated as a Boolean. If false, exceptions in constructing the SMCRA data structure are fatal. If true (DEFAULT), the exceptions are caught, but some residues or atoms will be missing. THESE EXCEPTIONS ARE DUE TO PROBLEMS IN THE PDB FILE!. - get_header - unused argument kept for historical compatibilty. - structure_builder - an optional user implemented StructureBuilder class. - QUIET - Evaluated as a Boolean. If true, warnings issued in constructing the SMCRA data will be suppressed. If false (DEFAULT), they will be shown. These warnings might be indicative of problems in the PDB file! """ # get_header is not used but is left in for API compatibility if structure_builder is not None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.trailer = None self.line_counter = 0 self.PERMISSIVE = bool(PERMISSIVE) self.QUIET = bool(QUIET) # Public methods def get_structure(self, id, file): """Return the structure. Arguments: - id - string, the id that will be used for the structure - file - name of the PDB file OR an open filehandle """ with warnings.catch_warnings(): if self.QUIET: warnings.filterwarnings("ignore", category=PDBConstructionWarning) self.header = None self.trailer = None # Make a StructureBuilder instance (pass id of structure as parameter) self.structure_builder.init_structure(id) with as_handle(file, mode="rU") as handle: lines = handle.readlines() if not lines: raise ValueError("Empty file.") self._parse(lines) self.structure_builder.set_header(self.header) # Return the Structure instance structure = self.structure_builder.get_structure() return structure def get_header(self): """Return the header.""" return self.header def get_trailer(self): """Return the trailer.""" return self.trailer # Private methods def _parse(self, header_coords_trailer): """Parse the PDB file (PRIVATE).""" # Extract the header; return the rest of the file self.header, coords_trailer = self._get_header(header_coords_trailer) # Parse the atomic data; return the PDB file trailer self.trailer = self._parse_coordinates(coords_trailer) def _get_header(self, header_coords_trailer): """Get the header of the PDB file, return the rest (PRIVATE).""" structure_builder = self.structure_builder i = 0 for i in range(0, len(header_coords_trailer)): structure_builder.set_line_counter(i + 1) line = header_coords_trailer[i] record_type = line[0:6] if record_type in ("ATOM ", "HETATM", "MODEL "): break header = header_coords_trailer[0:i] # Return the rest of the coords+trailer for further processing self.line_counter = i coords_trailer = header_coords_trailer[i:] header_dict = _parse_pdb_header_list(header) return header_dict, coords_trailer def _parse_coordinates(self, coords_trailer): """Parse the atomic data in the PDB file (PRIVATE).""" local_line_counter = 0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None for i in range(0, len(coords_trailer)): line = coords_trailer[i].rstrip("\n") record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 structure_builder.set_line_counter(global_line_counter) if record_type == "ATOM " or record_type == "HETATM": # Initialize the Model - there was no explicit MODEL record if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # atom name has internal spaces, e.g. " N B ", so # we do not strip spaces name = fullname else: # atom name is like " CA ", so we can strip spaces name = split_list[0] altloc = line[16] resname = line[17:20] chainid = line[21] try: serial_number = int(line[6:11]) except Exception: serial_number = 0 resseq = int(line[22:26].split()[0]) # sequence identifier icode = line[26] # insertion code if record_type == "HETATM": # hetero atom flag if resname == "HOH" or resname == "WAT": hetero_flag = "W" else: hetero_flag = "H" else: hetero_flag = " " residue_id = (hetero_flag, resseq, icode) # atomic coordinates try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except Exception: # Should we allow parsing to continue in permissive mode? # If so, what coordinates should we default to? Easier to abort! raise PDBConstructionException( "Invalid or missing coordinate(s) at line %i." % global_line_counter) coord = numpy.array((x, y, z), "f") # occupancy & B factor try: occupancy = float(line[54:60]) except Exception: self._handle_PDB_exception("Invalid or missing occupancy", global_line_counter) occupancy = None # Rather than arbitrary zero or one if occupancy is not None and occupancy < 0: # TODO - Should this be an error in strict mode? # self._handle_PDB_exception("Negative occupancy", # global_line_counter) # This uses fixed text so the warning occurs once only: warnings.warn( "Negative occupancy in one or more atoms", PDBConstructionWarning, ) try: bfactor = float(line[60:66]) except Exception: self._handle_PDB_exception("Invalid or missing B factor", global_line_counter) bfactor = 0.0 # PDB uses a default of zero if missing segid = line[72:76] element = line[76:78].strip().upper() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif current_residue_id != residue_id or current_resname != resname: current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) # init atom try: structure_builder.init_atom( name, coord, bfactor, occupancy, altloc, fullname, serial_number, element, ) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif record_type == "ANISOU": anisou = [ float(x) for x in ( line[28:35], line[35:42], line[43:49], line[49:56], line[56:63], line[63:70], ) ] # U's are scaled by 10^4 anisou_array = (numpy.array(anisou, "f") / 10000.0).astype("f") structure_builder.set_anisou(anisou_array) elif record_type == "MODEL ": try: serial_num = int(line[10:14]) except Exception: self._handle_PDB_exception( "Invalid or missing model serial number", global_line_counter) serial_num = 0 structure_builder.init_model(current_model_id, serial_num) current_model_id += 1 model_open = 1 current_chain_id = None current_residue_id = None elif record_type == "END " or record_type == "CONECT": # End of atomic data, return the trailer self.line_counter += local_line_counter return coords_trailer[local_line_counter:] elif record_type == "ENDMDL": model_open = 0 current_chain_id = None current_residue_id = None elif record_type == "SIGUIJ": # standard deviation of anisotropic B factor siguij = [ float(x) for x in ( line[28:35], line[35:42], line[42:49], line[49:56], line[56:63], line[63:70], ) ] # U sigma's are scaled by 10^4 siguij_array = (numpy.array(siguij, "f") / 10000.0).astype("f") structure_builder.set_siguij(siguij_array) elif record_type == "SIGATM": # standard deviation of atomic positions sigatm = [ float(x) for x in ( line[30:38], line[38:45], line[46:54], line[54:60], line[60:66], ) ] sigatm_array = numpy.array(sigatm, "f") structure_builder.set_sigatm(sigatm_array) local_line_counter += 1 # EOF (does not end in END or CONECT) self.line_counter = self.line_counter + local_line_counter return [] def _handle_PDB_exception(self, message, line_counter): """Handle exception (PRIVATE). This method catches an exception that occurs in the StructureBuilder object (if PERMISSIVE), or raises it again, this time adding the PDB line number to the error message. """ message = "%s at line %i." % (message, line_counter) if self.PERMISSIVE: # just print a warning - some residues/atoms may be missing warnings.warn( "PDBConstructionException: %s\n" "Exception ignored.\n" "Some atoms or residues may be missing in the data structure." % message, PDBConstructionWarning, ) else: # exceptions are fatal - raise again with new message (including line nr) raise PDBConstructionException(message)
class PDBParser: '''Parse a PDB file and return a Structure object.''' def __init__(self, structure_builder=None): # get_header is not used but is left in for API compatibility if structure_builder is not None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.line_counter = 0 # NOTE: Public methods def get_structure(self, id, file): '''Return the structure. Arguments: - id - string, the id that will be used for the structure - file - name of the PDB file OR an open filehandle ''' with warnings.catch_warnings(): # Make a StructureBuilder instance # (pass id of structure as parameter) self.structure_builder.init_structure(id) with as_handle(file) as handle: lines = [line.decode() for line in handle.readlines()] if not lines: raise PDBConstructionException('Empty file.') self._parse(lines) self.structure_builder.set_header(self.header) # Return the Structure instance # structure = self.structure_builder.get_structure() return # NOTE: Private methods def _parse(self, header_coords_trailer): '''Parse the PDB file (PRIVATE).''' # Parse the atomic data; return the PDB file trailer self.trailer = self._parse_coordinates(header_coords_trailer) def _parse_coordinates(self, coords_trailer): '''Parse the atomic data in the PDB file (PRIVATE).''' allowed_records = { 'ATOM ', 'HETATM', 'MODEL ', 'ENDMDL', 'TER ', 'ANISOU', # These are older 2.3 format specs: 'SIGATM', 'SIGUIJ', # bookkeeping records after coordinates: 'MASTER', # Additional records to not worry about: 'REMARK', 'JRNL ', 'LINK ', 'HELIX ', 'HETNAM', 'FORMUL', 'SOURCE', 'SEQRES', 'KEYWDS', 'SCALE1', 'SCALE2', 'SCALE3', 'SSBOND', 'HEADER', 'TITLE ', 'CRYST1', 'ORIGX1', 'ORIGX2', 'ORIGX3', 'SHEET ', 'HET ', 'COMPND', 'REVDAT', 'AUTHOR', 'EXPDTA', 'DBREF ', 'HETSYN', } local_line_counter = 0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None for i in range(0, len(coords_trailer)): line = coords_trailer[i].rstrip('\n') record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 structure_builder.set_line_counter(global_line_counter) if not line.strip(): continue # skip empty lines elif record_type.startswith('ATOM'): # Initialize the Model - there was no explicit MODEL record if len(line) > 80: raise PDBConstructionException( f'Record on line {i+1} of PBD is greater than 80 ' 'characters in length') if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] try: # check if starts with number int(fullname) except Exception: # if does not, continue pass else: raise PDBConstructionException( f'Atom name on line {i+1} of PDB is formated incorrectly. ' f'Received: "{fullname}". ' 'Expected an atom name beginning with a letter.') # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # atom name has internal spaces, e.g. ' N B ', so # we do not strip spaces name = fullname else: # atom name is like ' CA ', so we can strip spaces name = split_list[0] altloc = line[16] resname = line[17:20].strip() chainid = line[21] try: serial_number = int(line[6:11]) except Exception: serial_number = 0 try: resseq = int(line[22:26].split()[0]) # sequence identifier except Exception: raise PDBConstructionException( f'Residue sequence number on line {i+1} of PDB ' 'formated incorrectly. Received: ' f'{line[22:26].split()[0]}') icode = line[26] # insertion code hetero_flag = ' ' residue_id = (hetero_flag, resseq, icode) # atomic coordinates try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except Exception: raise PDBConstructionException( f'Invalid or missing coordinate(s) on line {i+1}. ' 'X, Y, and Z coordinates should be provided as floats. ' f'Received: X: {line[30:38]}, Y: {line[38:46]}, ' f'Z: {line[46:54]}.') coord = numpy.array((x, y, z), 'f') # occupancy & B factor try: occupancy = float(line[54:60]) except Exception: # Set the occupancy to 0.00 as a default if the provided # value is incorrect occupancy = 0.0 # NOTE: Uncomment to throw an exception for invalid # occupancy # raise PDBConstructionException( # f'Invalid or missing occupancy on line {i+1}. ' # 'The occupancy should be provided as a float value. ' # f'Received: {line[54:60]}.' # ) if occupancy is not None and occupancy < 0: raise PDBConstructionException( f'Negative occupancy on line {i+1}. ' 'The expected occupancy should be non-negative.') try: bfactor = float(line[60:66]) except Exception: # Set the B Factor to 0.00 as a default if the provided # value is incorrect bfactor = 0.0 # NOTE: Uncomment to throw an exception for invalid # B Factor # raise PDBConstructionException( # f'Invalid or missing B factor on line {i+1}. ' # 'The B factor should be provided as a float value. ' # f'Received: {line[60:66]}.' # ) segid = line[72:76] element = line[76:78].strip().upper() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: raise PDBConstructionException(message + f' on line {i+1}.') elif (current_residue_id != residue_id or current_resname != resname): current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: raise PDBConstructionException(message + f' on line {i+1}.') try: structure_builder.init_atom( name, coord, bfactor, occupancy, altloc, fullname, serial_number, element, ) except PDBConstructionException as message: raise PDBConstructionException(message + f' on line {i+1}.') elif (record_type.startswith('END') or record_type.startswith('CONECT') or record_type.startswith('ENDMDL')): # End of atomic data, return the trailer self.line_counter += local_line_counter return coords_trailer[local_line_counter:] elif record_type not in allowed_records: # NOTE: The following code enables Users to determine # whether an invalid record has been used # raise PDBConstructionException( # f'Ignoring unrecognized record "{record_type}" ' # f'at line {i+1}' # ) pass local_line_counter += 1 # EOF (does not end in END or CONECT) self.line_counter = self.line_counter + local_line_counter return []
class PDBParser(object): """Parse a PDB file and return a Structure object.""" def __init__(self, PERMISSIVE=True, get_header=False, structure_builder=None, QUIET=False): """Create a PDBParser object. The PDB parser call a number of standard methods in an aggregated StructureBuilder object. Normally this object is instanciated by the PDBParser object itself, but if the user provides his/her own StructureBuilder object, the latter is used instead. Arguments: - PERMISSIVE - Evaluated as a Boolean. If false, exceptions in constructing the SMCRA data structure are fatal. If true (DEFAULT), the exceptions are caught, but some residues or atoms will be missing. THESE EXCEPTIONS ARE DUE TO PROBLEMS IN THE PDB FILE!. - structure_builder - an optional user implemented StructureBuilder class. - QUIET - Evaluated as a Boolean. If true, warnings issued in constructing the SMCRA data will be suppressed. If false (DEFAULT), they will be shown. These warnings might be indicative of problems in the PDB file! """ if structure_builder is not None: self.structure_builder = structure_builder else: self.structure_builder = StructureBuilder() self.header = None self.trailer = None self.line_counter = 0 self.PERMISSIVE = bool(PERMISSIVE) self.QUIET = bool(QUIET) # Public methods def get_structure(self, id, file): """Return the structure. Arguments: - id - string, the id that will be used for the structure - file - name of the PDB file OR an open filehandle """ with warnings.catch_warnings(): if self.QUIET: warnings.filterwarnings("ignore", category=PDBConstructionWarning) self.header = None self.trailer = None # Make a StructureBuilder instance (pass id of structure as parameter) self.structure_builder.init_structure(id) with as_handle(file, mode='rU') as handle: self._parse(handle.readlines()) self.structure_builder.set_header(self.header) # Return the Structure instance structure = self.structure_builder.get_structure() return structure def get_header(self): """Return the header.""" return self.header def get_trailer(self): """Return the trailer.""" return self.trailer # Private methods def _parse(self, header_coords_trailer): """Parse the PDB file (PRIVATE).""" # Extract the header; return the rest of the file self.header, coords_trailer = self._get_header(header_coords_trailer) # Parse the atomic data; return the PDB file trailer self.trailer = self._parse_coordinates(coords_trailer) def _get_header(self, header_coords_trailer): """Get the header of the PDB file, return the rest (PRIVATE).""" structure_builder = self.structure_builder i = 0 for i in range(0, len(header_coords_trailer)): structure_builder.set_line_counter(i + 1) line = header_coords_trailer[i] record_type = line[0:6] if record_type == "ATOM " or record_type == "HETATM" or record_type == "MODEL ": break header = header_coords_trailer[0:i] # Return the rest of the coords+trailer for further processing self.line_counter = i coords_trailer = header_coords_trailer[i:] header_dict = _parse_pdb_header_list(header) return header_dict, coords_trailer def _parse_coordinates(self, coords_trailer): """Parse the atomic data in the PDB file (PRIVATE).""" local_line_counter = 0 structure_builder = self.structure_builder current_model_id = 0 # Flag we have an open model model_open = 0 current_chain_id = None current_segid = None current_residue_id = None current_resname = None for i in range(0, len(coords_trailer)): line = coords_trailer[i].rstrip('\n') record_type = line[0:6] global_line_counter = self.line_counter + local_line_counter + 1 structure_builder.set_line_counter(global_line_counter) if record_type == "ATOM " or record_type == "HETATM": # Initialize the Model - there was no explicit MODEL record if not model_open: structure_builder.init_model(current_model_id) current_model_id += 1 model_open = 1 fullname = line[12:16] # get rid of whitespace in atom names split_list = fullname.split() if len(split_list) != 1: # atom name has internal spaces, e.g. " N B ", so # we do not strip spaces name = fullname else: # atom name is like " CA ", so we can strip spaces name = split_list[0] altloc = line[16] resname = line[17:20] chainid = line[21] try: serial_number = int(line[6:11]) except Exception: serial_number = 0 resseq = int(line[22:26].split()[0]) # sequence identifier icode = line[26] # insertion code if record_type == "HETATM": # hetero atom flag if resname == "HOH" or resname == "WAT": hetero_flag = "W" else: hetero_flag = "H" else: hetero_flag = " " residue_id = (hetero_flag, resseq, icode) # atomic coordinates try: x = float(line[30:38]) y = float(line[38:46]) z = float(line[46:54]) except Exception: # Should we allow parsing to continue in permissive mode? # If so, what coordinates should we default to? Easier to abort! raise PDBConstructionException("Invalid or missing coordinate(s) at line %i." % global_line_counter) coord = numpy.array((x, y, z), "f") # occupancy & B factor try: occupancy = float(line[54:60]) except Exception: self._handle_PDB_exception("Invalid or missing occupancy", global_line_counter) occupancy = None # Rather than arbitrary zero or one if occupancy is not None and occupancy < 0: # TODO - Should this be an error in strict mode? # self._handle_PDB_exception("Negative occupancy", # global_line_counter) # This uses fixed text so the warning occurs once only: warnings.warn("Negative occupancy in one or more atoms", PDBConstructionWarning) try: bfactor = float(line[60:66]) except Exception: self._handle_PDB_exception("Invalid or missing B factor", global_line_counter) bfactor = 0.0 # The PDB use a default of zero if the data is missing segid = line[72:76] element = line[76:78].strip().upper() if current_segid != segid: current_segid = segid structure_builder.init_seg(current_segid) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif current_residue_id != residue_id or current_resname != resname: current_residue_id = residue_id current_resname = resname try: structure_builder.init_residue(resname, hetero_flag, resseq, icode) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) # init atom try: structure_builder.init_atom(name, coord, bfactor, occupancy, altloc, fullname, serial_number, element) except PDBConstructionException as message: self._handle_PDB_exception(message, global_line_counter) elif record_type == "ANISOU": anisou = [float(x) for x in (line[28:35], line[35:42], line[43:49], line[49:56], line[56:63], line[63:70])] # U's are scaled by 10^4 anisou_array = (numpy.array(anisou, "f") / 10000.0).astype("f") structure_builder.set_anisou(anisou_array) elif record_type == "MODEL ": try: serial_num = int(line[10:14]) except Exception: self._handle_PDB_exception("Invalid or missing model serial number", global_line_counter) serial_num = 0 structure_builder.init_model(current_model_id, serial_num) current_model_id += 1 model_open = 1 current_chain_id = None current_residue_id = None elif record_type == "END " or record_type == "CONECT": # End of atomic data, return the trailer self.line_counter += local_line_counter return coords_trailer[local_line_counter:] elif record_type == "ENDMDL": model_open = 0 current_chain_id = None current_residue_id = None elif record_type == "SIGUIJ": # standard deviation of anisotropic B factor siguij = [float(x) for x in (line[28:35], line[35:42], line[42:49], line[49:56], line[56:63], line[63:70])] # U sigma's are scaled by 10^4 siguij_array = (numpy.array(siguij, "f") / 10000.0).astype("f") structure_builder.set_siguij(siguij_array) elif record_type == "SIGATM": # standard deviation of atomic positions sigatm = [float(x) for x in (line[30:38], line[38:45], line[46:54], line[54:60], line[60:66])] sigatm_array = numpy.array(sigatm, "f") structure_builder.set_sigatm(sigatm_array) local_line_counter += 1 # EOF (does not end in END or CONECT) self.line_counter = self.line_counter + local_line_counter return [] def _handle_PDB_exception(self, message, line_counter): """Handle exception (PRIVATE). This method catches an exception that occurs in the StructureBuilder object (if PERMISSIVE), or raises it again, this time adding the PDB line number to the error message. """ message = "%s at line %i." % (message, line_counter) if self.PERMISSIVE: # just print a warning - some residues/atoms may be missing warnings.warn("PDBConstructionException: %s\n" "Exception ignored.\n" "Some atoms or residues may be missing in the data structure." % message, PDBConstructionWarning) else: # exceptions are fatal - raise again with new message (including line nr) raise PDBConstructionException(message)
def pdb_extract(structure, **kwargs): # model to extract from pdb extract_model = None if not 'model' in kwargs else kwargs['model'] new_model_id = -1 if not 'new_model' in kwargs else kwargs['new_model'] extract_chain = None if not 'chain' in kwargs else kwargs['chain'] first_res = None if not 'first_res' in kwargs else kwargs['first_res'] last_res = None if not 'last_res' in kwargs else kwargs['last_res'] new_first_res = None if not 'new_first_res' in kwargs else kwargs[ 'new_first_res'] gap_count = 0 if not 'gap_count' in kwargs else kwargs['gap_count'] water_id = None if not 'water' in kwargs else kwargs['water'] model_rebumber_flag = bool((extract_model is not None) or new_model_id >= 0) res_renumber_flag = bool(first_res or last_res or new_first_res or gap_count) structure_builder = StructureBuilder() structure_builder.init_structure('pdb_extract') structure_builder.set_line_counter(0) line_counter = 0 start_resseq_by_default = 1 if not new_first_res else new_first_res for model in structure: if model_rebumber_flag and \ ( extract_model is not None ) and model.get_id() != extract_model: continue if model_rebumber_flag and new_model_id >= 0: this_model_id = new_model_id new_model_id += 1 else: this_model_id = model.get_id() structure_builder.init_model(this_model_id, this_model_id) for chain in model: if extract_chain and chain.get_id() != extract_chain: continue structure_builder.init_seg(' ') structure_builder.init_chain(chain.get_id()) resdict = {} if res_renumber_flag: # first_res = res_range_tuple[0] # last_res = res_range_tuple[1] resdict['before'] = select_residues_from_chain( chain, first_res=first_res, gap_count=gap_count) resdict['hit'] = select_residues_from_chain( chain, first_res=first_res, last_res=last_res) resdict['after'] = select_residues_from_chain( chain, last_res=last_res, gap_count=gap_count) else: resdict['before'] = [] resdict['hit'] = chain.get_list() resdict['after'] = [] new_resseq = start_resseq_by_default - len(resdict['before']) resdict['water'] = chain_water_id(chain, water_id) for key in ['before', 'hit', 'after', 'water']: for residue in resdict[key]: if res_renumber_flag: new_resid = ' ', new_resseq, ' ' else: new_resid = residue.get_id() structure_builder.init_residue(residue.get_resname(), *new_resid) residue_atoms = None if key == 'before': residue_atoms = [atom for atom in residue if \ (atom.get_name() == 'C' or atom.get_name() == 'O')] elif key == 'hit' or key == 'water': residue_atoms = residue.get_list() elif key == 'after': residue_atoms = [atom for atom in residue if \ (atom.get_name() == 'N' or atom.get_name() == 'HN')] for atom in residue_atoms: structure_builder.init_atom(atom.get_name(), atom.get_coord(), atom.get_bfactor(), atom.get_occupancy(), atom.get_altloc(), atom.get_fullname()) structure_builder.set_line_counter(line_counter) line_counter += 1 new_resseq += 1 if key == 'water' and gap_count and len( resdict['after']) != gap_count: new_resseq += gap_count - len(resdict['after']) out_structure = structure_builder.get_structure() return out_structure