Check PDB Symmetry

Determine whether your atomic model has been built into the unit cell with the highest possible space group symmetry.  This validation can be done at any time, with either a PDB-formatted model, or with a combination of PDB model and structure factors.  

Program Documentation

1) find potential symmetry for a coordinate file & optionally an mtz file
labelit.check_pdb_symmetry <PDB coordinate file> [data=<mtz file>]

2) find potential symmetry in a Data Bank entry, given the 4-character PDB code
labelit.check_pdb_symmetry <pdb code>
(this feature is not quite implemented yet but is promised soon)

Other command line arguments available
labelit.check_pdb_symmetry --help

Important:  Each command line argument must be typed without spaces, as shown here:

 (no keyword) Name of PDB file
 data=None Name of MTZ file
 asu_file=None|<base> If given, this switches on file output and represents the base name for the output pdb files.  Three types of files are written:
  1. For each coset N, an asymmetric unit model in the target (higher) symmetry space group (containing only the matched peptides), to file <base>_coset<N>.pdb.
  2. For each coset N, a degenerate model in the input space group containing all atoms in the original structure but with potential symmetry applied, to file <base>_degen<N>.pdb.
  3. A structure factor file containing observations merged into the higher symmetry space group.  File <base>_amplitude_<sg>.mtz (<sg> is the higher symmetry group), data label FOBS_MERGE_EQ.
    • Only use this file if the original diffraction images are not available, otherwise re-integrate the data into the higher space group, using labelit.index pdb_file=<base>_coset0.pdb to constrain to the correct symmetry.
    • If the input column label "R-free-flags" exists, then Rfree flags will be merged into the higher-symmetry space group, marking any reflection as rfree=1 if any one of its lower-symmetry equivalents was set to 1.
 asu_file_scratch=False If True, an additional file is written, <base>_scratch<N>.pdb, giving all input atoms,
transformed in to the target (higher) symmetry space group setting.  This is intended chiefly for transforming ligand atoms, which are not assigned to coset asu models by the program.
 d_min=-1.0 Override for the outer resolution cutoff (Angstroms)
 ignore_chain=None For a structure to qualify as having higher symmetry, each chain must be matched.  Use ignore_chain=A to remove chain 'A' from the matching requirement, if it is known ahead of time to lack a higher-symmetry match.  This is useful for oligopeptides.  Program does not currently permit multiple chain id's (contact authors for fix).
 max_delta=3.0 Maximum Le Page delta (degrees) for determining the lattice symmetry 
 r_symop_tol=0.25   Maximum allowable symmetry operator R factor on calculated intensities (eq 1)
 peak_cutoff=0.8      Minimum height of peak from fast translation search (eq 2)
 score_tol=0.25        Maximum allowable phase weighted score (eq 3)
 rotation_tolerance=0.1 Tolerance for rotation match of two alpha-carbon sets (fractional coords)
 translation_tolerance=0.1 Tolerance for translational match of two alpha-carbon sets (fractional)
 verbose_match=False If True, amino acid sequences for the potentially symmetry-matched chains are printed.

Current Limitations
  1. The program treats polypeptides only.  Nucleotides and non-peptide ligands have not been implemented.
  2. The program tests rotational symmetry only.  It is also possible to test for potentially higher symmetry using translational symmetry operators (for example, by finding strong Patterson peaks at rational fraction positions).  This will be implemented in the future.

Nicholas Sauter,
Mar 23, 2010, 12:57 PM