Module wwpdb

source code

PDB structure parsers, format builders and database providers.

The most basic usage is:

>>> parser = StructureParser('structure.pdb')
>>> parser.parse_structure()
<Structure>     # a Structure object (model)

or if this is an NMR ensemble:

>>> parser.parse_models()
<Ensemble>      # an Ensemble object (collection of alternative Structure-s)

This module introduces a family of PDB file parsers. The common interface of all parsers is defined in AbstractStructureParser. This class has several implementations:

• RegularStructureParser - handles normal PDB files with SEQRES fields
• LegacyStructureParser - reads structures from legacy or malformed PDB files, which are lacking SEQRES records (initializes all residues from the ATOMs instead)
• PDBHeaderParser - reads only the headers of the PDB files and produces structures without coordinates. Useful for reading metadata (e.g. accession numbers or just plain SEQRES sequences) with minimum overhead

Unless you have a special reason, you should use the StructureParser factory, which returns a proper AbstractStructureParser implementation, depending on the input PDB file. If the input file looks like a regular PDB file, the factory returns a RegularStructureParser, otherwise it instantiates LegacyStructureParser. StructureParser is in fact an alias for AbstractStructureParser.create_parser.

Writing your own, customized PDB parser is easy. Suppose that you are trying to parse a PDB-like file which misuses the charge column to store custom info. This will certainly crash RegularStructureParser (for good), but you can create your own parser as a workaround. All you need to to is to override the virtual _read_charge hook method:

   class CustomParser(RegularStructureParser):
   
       def _read_charge(self, line):
           try:
               return super(CustomParser, self)._read_charge(line)
           except StructureFormatError:
               return None

Another important abstraction in this module is StructureProvider. It has several implementations which can be used to retrieve PDB Structures from various sources: file system directories, remote URLs, etc. You can easily create your own provider as well. See StructureProvider for details.

Finally, this module gives you some FileBuilders, used for text serialization of Structures and Ensembles:

>>> builder = PDBFileBuilder(stream)
>>> builder.add_header(structure)
>>> builder.add_structure(structure)

where stream is any Python stream, e.g. an open file or sys.stdout.

See Ensemble and Structure from csb.bio.structure for details on these objects.