# Copyright (C) 2002, Thomas Hamelryck (thamelry@binf.ku.dk) # This code is part of the Biopython distribution and governed by its # license. Please see the LICENSE file that should have been included # as part of this package. """mmCIF parsers.""" import warnings import numpy as np from Bio.File import as_handle from Bio.PDB.MMCIF2Dict import MMCIF2Dict from Bio.PDB.PDBExceptions import PDBConstructionException from Bio.PDB.PDBExceptions import PDBConstructionWarning from Bio.PDB.StructureBuilder import StructureBuilder class MMCIFParser: """Parse a mmCIF file and return a Structure object.""" def __init__( self, structure_builder=None, auth_chains=True, auth_residues=True, QUIET=False ): """Create a PDBParser object. The mmCIF parser calls a number of standard methods in an aggregated StructureBuilder object. Normally this object is instantiated by the MMCIParser object itself, but if the user provides his/her own StructureBuilder object, the latter is used instead. Arguments: - structure_builder - an optional user implemented StructureBuilder class. - auth_chains - True by default. If true, use the author chain IDs. If false, use the re-assigned mmCIF chain IDs. - auth_residues - True by default. If true, use the author residue numbering. If false, use the mmCIF "label" residue numbering, which has no insertion codes, and strictly increments residue numbers. NOTE: Non-polymers such as water don't have a "label" residue number, and will be skipped. - 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 mmCIF 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.build_structure = None self.auth_chains = bool(auth_chains) self.auth_residues = bool(auth_residues) self.QUIET = bool(QUIET) # Public methods def get_structure(self, structure_id, filename): """Return the structure. Arguments: - structure_id - string, the id that will be used for the structure - filename - name of mmCIF file, OR an open text mode file handle """ with warnings.catch_warnings(): if self.QUIET: warnings.filterwarnings("ignore", category=PDBConstructionWarning) self._mmcif_dict = MMCIF2Dict(filename) self._build_structure(structure_id) self._structure_builder.set_header(self._get_header()) return self._structure_builder.get_structure() # Private methods def _mmcif_get(self, key, dict, deflt): if key in dict: rslt = dict[key][0] if "?" != rslt: return rslt return deflt def _update_header_entry(self, target_key, keys): md = self._mmcif_dict for key in keys: val = md.get(key) try: item = val[0] except (TypeError, IndexError): continue if item != "?" and item != ".": self.header[target_key] = item break def _get_header(self): self.header = { "name": "", "head": "", "idcode": "", "deposition_date": "", "structure_method": "", "resolution": None, } self._update_header_entry( "idcode", ["_entry_id", "_exptl.entry_id", "_struct.entry_id"] ) self._update_header_entry("name", ["_struct.title"]) self._update_header_entry( "head", ["_struct_keywords.pdbx_keywords", "_struct_keywords.text"] ) self._update_header_entry( "deposition_date", ["_pdbx_database_status.recvd_initial_deposition_date"] ) self._update_header_entry("structure_method", ["_exptl.method"]) self._update_header_entry( "resolution", [ "_refine.ls_d_res_high", "_refine_hist.d_res_high", "_em_3d_reconstruction.resolution", ], ) if self.header["resolution"] is not None: try: self.header["resolution"] = float(self.header["resolution"]) except ValueError: self.header["resolution"] = None return self.header def _build_structure(self, structure_id): # two special chars as placeholders in the mmCIF format # for item values that cannot be explicitly assigned # see: pdbx/mmcif syntax web page _unassigned = {".", "?"} mmcif_dict = self._mmcif_dict atom_serial_list = mmcif_dict["_atom_site.id"] atom_id_list = mmcif_dict["_atom_site.label_atom_id"] residue_id_list = mmcif_dict["_atom_site.label_comp_id"] try: element_list = mmcif_dict["_atom_site.type_symbol"] except KeyError: element_list = None if self.auth_chains: chain_id_list = mmcif_dict["_atom_site.auth_asym_id"] else: chain_id_list = mmcif_dict["_atom_site.label_asym_id"] x_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_x"]] y_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_y"]] z_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_z"]] alt_list = mmcif_dict["_atom_site.label_alt_id"] icode_list = mmcif_dict["_atom_site.pdbx_PDB_ins_code"] b_factor_list = mmcif_dict["_atom_site.B_iso_or_equiv"] occupancy_list = mmcif_dict["_atom_site.occupancy"] fieldname_list = mmcif_dict["_atom_site.group_PDB"] try: serial_list = [int(n) for n in mmcif_dict["_atom_site.pdbx_PDB_model_num"]] except KeyError: # No model number column serial_list = None except ValueError: # Invalid model number (malformed file) raise PDBConstructionException("Invalid model number") from None try: aniso_u11 = mmcif_dict["_atom_site_anisotrop.U[1][1]"] aniso_u12 = mmcif_dict["_atom_site_anisotrop.U[1][2]"] aniso_u13 = mmcif_dict["_atom_site_anisotrop.U[1][3]"] aniso_u22 = mmcif_dict["_atom_site_anisotrop.U[2][2]"] aniso_u23 = mmcif_dict["_atom_site_anisotrop.U[2][3]"] aniso_u33 = mmcif_dict["_atom_site_anisotrop.U[3][3]"] aniso_flag = 1 except KeyError: # no anisotropic B factors aniso_flag = 0 if self.auth_residues: # if auth_seq_id is present, we use this. # Otherwise label_seq_id is used. if "_atom_site.auth_seq_id" in mmcif_dict: seq_id_list = mmcif_dict["_atom_site.auth_seq_id"] else: seq_id_list = mmcif_dict["_atom_site.label_seq_id"] else: seq_id_list = mmcif_dict["_atom_site.label_seq_id"] # Now loop over atoms and build the structure current_chain_id = None current_residue_id = None current_resname = None structure_builder = self._structure_builder structure_builder.init_structure(structure_id) structure_builder.init_seg(" ") # Historically, Biopython PDB parser uses model_id to mean array index # so serial_id means the Model ID specified in the file current_model_id = -1 current_serial_id = -1 for i in range(len(atom_id_list)): # set the line_counter for 'ATOM' lines only and not # as a global line counter found in the PDBParser() structure_builder.set_line_counter(i) # Try coercing serial to int, for compatibility with PDBParser # But do not quit if it fails. mmCIF format specs allow strings. try: serial = int(atom_serial_list[i]) except ValueError: serial = atom_serial_list[i] warnings.warn( "PDBConstructionWarning: Some atom serial numbers are not numerical", PDBConstructionWarning, ) x = x_list[i] y = y_list[i] z = z_list[i] resname = residue_id_list[i] chainid = chain_id_list[i] altloc = alt_list[i] if altloc in _unassigned: altloc = " " resseq = seq_id_list[i] if resseq == ".": # Non-existing residue ID try: msg_resseq = mmcif_dict["_atom_site.auth_seq_id"][i] msg = f"Non-existing residue ID in chain '{chainid}', residue '{msg_resseq}'" except (KeyError, IndexError): msg = f"Non-existing residue ID in chain '{chainid}'" warnings.warn( "PDBConstructionWarning: " + msg, PDBConstructionWarning, ) continue int_resseq = int(resseq) icode = icode_list[i] if icode in _unassigned: icode = " " name = atom_id_list[i] # occupancy & B factor try: tempfactor = float(b_factor_list[i]) except ValueError: raise PDBConstructionException("Invalid or missing B factor") from None try: occupancy = float(occupancy_list[i]) except ValueError: raise PDBConstructionException("Invalid or missing occupancy") from None fieldname = fieldname_list[i] if fieldname == "HETATM": if resname == "HOH" or resname == "WAT": hetatm_flag = "W" else: hetatm_flag = "H" else: hetatm_flag = " " resseq = (hetatm_flag, int_resseq, icode) if serial_list is not None: # model column exists; use it serial_id = serial_list[i] if current_serial_id != serial_id: # if serial changes, update it and start new model current_serial_id = serial_id current_model_id += 1 structure_builder.init_model(current_model_id, current_serial_id) current_chain_id = None current_residue_id = None current_resname = None else: # no explicit model column; initialize single model structure_builder.init_model(current_model_id) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = None current_resname = None if current_residue_id != resseq or current_resname != resname: current_residue_id = resseq current_resname = resname structure_builder.init_residue(resname, hetatm_flag, int_resseq, icode) coord = np.array((x, y, z), "f") element = element_list[i].upper() if element_list else None structure_builder.init_atom( name, coord, tempfactor, occupancy, altloc, name, serial_number=serial, element=element, ) if aniso_flag == 1 and i < len(aniso_u11): u = ( aniso_u11[i], aniso_u12[i], aniso_u13[i], aniso_u22[i], aniso_u23[i], aniso_u33[i], ) mapped_anisou = [float(_) for _ in u] anisou_array = np.array(mapped_anisou, "f") structure_builder.set_anisou(anisou_array) # Now try to set the cell try: a = float(mmcif_dict["_cell.length_a"][0]) b = float(mmcif_dict["_cell.length_b"][0]) c = float(mmcif_dict["_cell.length_c"][0]) alpha = float(mmcif_dict["_cell.angle_alpha"][0]) beta = float(mmcif_dict["_cell.angle_beta"][0]) gamma = float(mmcif_dict["_cell.angle_gamma"][0]) cell = np.array((a, b, c, alpha, beta, gamma), "f") spacegroup = mmcif_dict["_symmetry.space_group_name_H-M"][0] spacegroup = spacegroup[1:-1] # get rid of quotes!! if spacegroup is None: raise Exception structure_builder.set_symmetry(spacegroup, cell) except Exception: pass # no cell found, so just ignore class FastMMCIFParser: """Parse an MMCIF file and return a Structure object.""" def __init__( self, structure_builder=None, auth_chains=True, auth_residues=True, QUIET=False ): """Create a FastMMCIFParser object. The mmCIF parser calls a number of standard methods in an aggregated StructureBuilder object. Normally this object is instantiated by the parser object itself, but if the user provides his/her own StructureBuilder object, the latter is used instead. The main difference between this class and the regular MMCIFParser is that only 'ATOM' and 'HETATM' lines are parsed here. Use if you are interested only in coordinate information. Arguments: - structure_builder - an optional user implemented StructureBuilder class. - auth_chains - True by default. If true, use the author chain IDs. If false, use the re-assigned mmCIF chain IDs. - auth_residues - True by default. If true, use the author residue numbering. If false, use the mmCIF "label" residue numbering, which has no insertion codes, and strictly increments residue numbers. NOTE: Non-polymers such as water don't have a "label" residue number, and will be skipped. - 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 mmCIF file! """ if structure_builder is not None: self._structure_builder = structure_builder else: self._structure_builder = StructureBuilder() self.line_counter = 0 self.build_structure = None self.auth_chains = bool(auth_chains) self.auth_residues = bool(auth_residues) self.QUIET = bool(QUIET) # Public methods def get_structure(self, structure_id, filename): """Return the structure. Arguments: - structure_id - string, the id that will be used for the structure - filename - name of the mmCIF file OR an open filehandle """ with warnings.catch_warnings(): if self.QUIET: warnings.filterwarnings("ignore", category=PDBConstructionWarning) with as_handle(filename) as handle: self._build_structure(structure_id, handle) return self._structure_builder.get_structure() # Private methods def _build_structure(self, structure_id, filehandle): # two special chars as placeholders in the mmCIF format # for item values that cannot be explicitly assigned # see: pdbx/mmcif syntax web page _unassigned = {".", "?"} # Read only _atom_site. and atom_site_anisotrop entries read_atom, read_aniso = False, False _fields, _records = [], [] _anisof, _anisors = [], [] for line in filehandle: if line.startswith("_atom_site."): read_atom = True _fields.append(line.strip()) elif line.startswith("_atom_site_anisotrop."): read_aniso = True _anisof.append(line.strip()) elif read_atom and line.startswith("#"): read_atom = False elif read_aniso and line.startswith("#"): read_aniso = False elif read_atom: _records.append(line.strip()) elif read_aniso: _anisors.append(line.strip()) # Dumping the shlex module here since this particular # category should be rather straightforward. # Quite a performance boost.. _record_tbl = zip(*map(str.split, _records)) _anisob_tbl = zip(*map(str.split, _anisors)) mmcif_dict = dict(zip(_fields, _record_tbl)) mmcif_dict.update(dict(zip(_anisof, _anisob_tbl))) # Build structure object atom_serial_list = mmcif_dict["_atom_site.id"] atom_id_list = mmcif_dict["_atom_site.label_atom_id"] residue_id_list = mmcif_dict["_atom_site.label_comp_id"] try: element_list = mmcif_dict["_atom_site.type_symbol"] except KeyError: element_list = None if self.auth_chains: chain_id_list = mmcif_dict["_atom_site.auth_asym_id"] else: chain_id_list = mmcif_dict["_atom_site.label_asym_id"] x_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_x"]] y_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_y"]] z_list = [float(x) for x in mmcif_dict["_atom_site.Cartn_z"]] alt_list = mmcif_dict["_atom_site.label_alt_id"] icode_list = mmcif_dict["_atom_site.pdbx_PDB_ins_code"] b_factor_list = mmcif_dict["_atom_site.B_iso_or_equiv"] occupancy_list = mmcif_dict["_atom_site.occupancy"] fieldname_list = mmcif_dict["_atom_site.group_PDB"] try: serial_list = [int(n) for n in mmcif_dict["_atom_site.pdbx_PDB_model_num"]] except KeyError: # No model number column serial_list = None except ValueError: # Invalid model number (malformed file) raise PDBConstructionException("Invalid model number") from None try: aniso_u11 = mmcif_dict["_atom_site_anisotrop.U[1][1]"] aniso_u12 = mmcif_dict["_atom_site_anisotrop.U[1][2]"] aniso_u13 = mmcif_dict["_atom_site_anisotrop.U[1][3]"] aniso_u22 = mmcif_dict["_atom_site_anisotrop.U[2][2]"] aniso_u23 = mmcif_dict["_atom_site_anisotrop.U[2][3]"] aniso_u33 = mmcif_dict["_atom_site_anisotrop.U[3][3]"] aniso_flag = 1 except KeyError: # no anisotropic B factors aniso_flag = 0 if self.auth_residues: # if auth_seq_id is present, we use this. # Otherwise label_seq_id is used. if "_atom_site.auth_seq_id" in mmcif_dict: seq_id_list = mmcif_dict["_atom_site.auth_seq_id"] else: seq_id_list = mmcif_dict["_atom_site.label_seq_id"] else: seq_id_list = mmcif_dict["_atom_site.label_seq_id"] # Now loop over atoms and build the structure current_chain_id = None current_residue_id = None current_resname = None structure_builder = self._structure_builder structure_builder.init_structure(structure_id) structure_builder.init_seg(" ") # Historically, Biopython PDB parser uses model_id to mean array index # so serial_id means the Model ID specified in the file current_model_id = -1 current_serial_id = -1 for i in range(len(atom_id_list)): # set the line_counter for 'ATOM' lines only and not # as a global line counter found in the PDBParser() structure_builder.set_line_counter(i) serial = atom_serial_list[i] x = x_list[i] y = y_list[i] z = z_list[i] resname = residue_id_list[i] chainid = chain_id_list[i] altloc = alt_list[i] if altloc in _unassigned: altloc = " " resseq = seq_id_list[i] if resseq == ".": # Non-existing residue ID try: msg_resseq = mmcif_dict["_atom_site.auth_seq_id"][i] msg = f"Non-existing residue ID in chain '{chainid}', residue '{msg_resseq}'" except (KeyError, IndexError): msg = f"Non-existing residue ID in chain '{chainid}'" warnings.warn( "PDBConstructionWarning: " + msg, PDBConstructionWarning, ) continue int_resseq = int(resseq) icode = icode_list[i] if icode in _unassigned: icode = " " # Remove occasional " from quoted atom names (e.g. xNA) name = atom_id_list[i].strip('"') # occupancy & B factor try: tempfactor = float(b_factor_list[i]) except ValueError: raise PDBConstructionException("Invalid or missing B factor") from None try: occupancy = float(occupancy_list[i]) except ValueError: raise PDBConstructionException("Invalid or missing occupancy") from None fieldname = fieldname_list[i] if fieldname == "HETATM": hetatm_flag = "H" else: hetatm_flag = " " resseq = (hetatm_flag, int_resseq, icode) if serial_list is not None: # model column exists; use it serial_id = serial_list[i] if current_serial_id != serial_id: # if serial changes, update it and start new model current_serial_id = serial_id current_model_id += 1 structure_builder.init_model(current_model_id, current_serial_id) current_chain_id = None current_residue_id = None current_resname = None else: # no explicit model column; initialize single model structure_builder.init_model(current_model_id) if current_chain_id != chainid: current_chain_id = chainid structure_builder.init_chain(current_chain_id) current_residue_id = None current_resname = None if current_residue_id != resseq or current_resname != resname: current_residue_id = resseq current_resname = resname structure_builder.init_residue(resname, hetatm_flag, int_resseq, icode) coord = np.array((x, y, z), "f") element = element_list[i] if element_list else None structure_builder.init_atom( name, coord, tempfactor, occupancy, altloc, name, serial_number=serial, element=element, ) if aniso_flag == 1 and i < len(aniso_u11): u = ( aniso_u11[i], aniso_u12[i], aniso_u13[i], aniso_u22[i], aniso_u23[i], aniso_u33[i], ) mapped_anisou = [float(_) for _ in u] anisou_array = np.array(mapped_anisou, "f") structure_builder.set_anisou(anisou_array)