Challenges in visualization of complexes and the PDB

1
Number of released entries
Year
2
Growth of Molecular Complexity
Number of Structures Containing that Number of
Chains
Number of Chains
Year
3
Challenges in visualization of complexes and the
PDB
HEADER COMPLEX (ACETYLATION/ACTIN-BINDING)
30-MAY-97 1HLU TITLE
STRUCTURE OF BOVINE BETA-ACTIN-PROFILIN COMPLEX
WITH ACTIN TITLE 2 BOUND ATP
PHOSPHATES SOLVENT ACCESSIBLE
COMPND MOL_ID 1
COMPND
2 MOLECULE BETA-ACTIN
COMPND 3 CHAIN A

COMPND 4 MOL_ID 2
COMPND 5
MOLECULE PROFILIN
COMPND 6 CHAIN P

SOURCE MOL_ID 1
SOURCE 2
ORGANISM_SCIENTIFIC BOS TAURUS
SOURCE 3 ORGANISM_COMMON
BOVINE
SOURCE 4 ORGAN THYMUS
SOURCE 5
MOL_ID 2
SOURCE 6 ORGANISM_SCIENTIFIC
BOS TAURUS
SOURCE 7 ORGANISM_COMMON BOVINE
SOURCE 8 ORGAN
THYMUS
KEYWDS COMPLEX (ACETYLATION/ACTIN-
BINDING), ACTIN, PROFILIN,
KEYWDS 2 CONFORMATIONAL CHANGES, CYTOSKELETON
EXPDTA X-RAY
DIFFRACTION
AUTHOR J.K.CHIK,U.LINDBERG,C.E.S
CHUTT
REVDAT 1 15-OCT-97 1HLU 0
JRNL AUTH
J.K.CHIK,U.LINDBERG,C.E.SCHUTT
JRNL TITL THE STRUCTURE OF
AN OPEN STATE OF BETA-ACTIN AT JRNL
TITL 2 2.65 A RESOLUTION
JRNL REF
J.MOL.BIOL. V. 263 607 1996
JRNL REFN ASTM JMOBAK UK
ISSN 0022-2836 0070

• Visualization workshop, October 2003

4
What is PDBs role in molecular visualization?
Visualization software
Coordinate files of molecules
Visualization of molecules


5
What is PDBs responsibility with respect to
molecular visualization?
providing links explanations for visualization
software
providing complete correctly annotated
coordinate files in multiple formats (PDB, cif,
XML)
visualization for user community general users,
researchers, annotators, students, educators,
databases, bio-informaticians


6
The PDB format
JRNL TITL COMPARISON OF THE
THREE-DIMENSIONAL STRUCTURES OF JRNL
TITL 2 RECOMBINANT HUMAN H AND HORSE L
FERRITINS AT HIGH JRNL TITL 3
RESOLUTION
JRNL REF J.MOL.BIOL.
V. 268 424 1997 JRNL
REFN ASTM JMOBAK UK ISSN 0022-2836
0070
The mmcif format
_citation.id primary
_citation.title Comparison
of the three-dimensional structures of
recombinant human H and horse L ferritins at high
resolution. _citation.journal_abbrev
J.Mol.Biol. _citation.journal_volume
268 _citation.page_first 424
_citation.page_last 448
_citation.year 1997
_citation.journal_id_ASTM JMOBAK
_citation.country UK
_citation.journal_id_ISSN 0022-2836
_citation.journal_id_CSD 0070
_citation.book_publisher ?
_citation.pdbx_database_id_PubMed
9159481
mmCIF macromolecular Crystallographic
Information File This is an extension of the
Crystallographic Information File (CIF) data
representation (used for describing small
molecule structures) to describe macromolecules.
7
Complete and correctly annotated coordinate files

• Do PDB files conform to uniform standards?
• Yes, Remediated mmcif files are available. They
  can be converted to PDB format using CIFTr
  (This application is available at
  http//deposit.pdb.org/software/)
• Do PDB files contain coordinates for the complete
  biological unit?
• Yes, both coordinates and pictures of the
  biological unit of all PDB files are now
  available.

8
What is a biological unit?

• PDB has coordinates of molecules determined by
• X-ray crystallography
• NMR
• Electron microscopy
• Theoretical modeling
• Primary coordinate files for crystallographic
  structure generally contain one asymmetric
  (unique) unit.
• The biological molecule (also called a biological
  unit) is the macromolecule that has been shown to
  be or is believed to be functional. This could
  include one, a part of or multiple asymmetric
  units.

9
Concept of the biological unit
Biological unit
Biological unit could include one, a part of or
multiple asymmetric units.
10
Downloading biological unit images/ coordinate
files from the PDB
Information for constructing the biological unit
is contained in remark 350 of the PDB file
PDB ID 1AEW
11
Visualizing the biological unit
Some visualization tools fail to duplicate the
secondary structure records for symmetry related
molecules in the biological unit
Biological unit of 1AEW Viewed in RasMol
12
Large macromolecular assembly 1 Viruses

• For viruses, usually the coordinates of the
  icosohedral asymmetric unit are deposited to the
  PDB. Transformation matrices for generating the
  complete virus are also provided.
• Sometimes additional matrices are provided to
  generate the icosohedral asymmetric unit from the
  given coordinates

Virus particles have high symmetry (5, 3, 2)
13
Generating the biological unit of a virus
Coordinates in crystallographic Symmetry axes
Coordinates in icosohedral Symmetry axes
Conversion matrix
Asymmetric unit NCS applied
Asymmetric unit NCS not applied
Asymmetric unit
60 matrices
Recipe matrices
Crystallographic Symmetry operations
Biological unit
14
Virus problems and solutions
Problems - matrices for generating the
biological unit? - improper nomenclature of
the 60 matrices for generating the icosohedral
virus particle? - placeholder for the NCS
matrices for completing the icosohedral
asymmetric unit ? - conversion matrix between
crystallographic and icosohedral
axes? Possible solutions - uniform
representation of matrices for generating the
biological unit? - change in the nomenclature
of the 60 matrices for generating the
virus? - conversion matrix between
crystallographic and icosohedral axes always
available? - other?
15
Large macromolecular assembly 2 Ribosomes

• The current PDB format can hold
• a maximum of 99,999 atom records, and
• upto 62 different polymer chains.
• Since there is no way to represent structures
  that exceed either of these restrictions in a
  single PDB file we have divided such structures
  into multiple PDB entries.
• Although this is not a perfect solution, we have
  done this to support existing software that rely
  on current format.
• The mmCIF/ XML format has no such restrictions.

16
Ribosomes
1GIY Large subunit of the ribosome
1GIX Small subunit of the ribosome
17
Ribosome problems and solutions
Problems restrictions in the PDB file
format? Size of file? Scaling? Dockin
g? Visualization of nucleic acids?
Possible solutions use of mmCIF/ XML
format? better way to represent nucleic
acids? other?
18
What is PDBs responsibility with respect to
molecular visualization?
providing links explanations for visualization
software
providing complete correctly annotated
coordinate files in multiple formats (PDB, cif,
XML)
Visualization for user community general users,
researchers, annotators, students, educators,
databases, bio-informaticians


19
Annotator needs

• Currently we use RasMol and NdbView for quick
  visualization and checking
• For annotation, the visualization software should
  be capable of
• Quick display (especially for large complexes)
• Displaying secondary structure
• Selecting atoms or residues for display or
  rendering
• Showing symmetry related molecules
• Coloring all or selected residues or chains
• Computing distances
• Displaying standard and unusual ligands

20
Educator needs

• Visualization programs for educators and students
  should
• Be free
• Be open source
• Be capable of running on multiple platforms
  (without browser dependence)
• Be portable
• Be easy to install and use
• Have user friendly interface
• Common themes from a survey of 25 educators from
  all over the world

21
Educator requests and visions

• Visualization software for educators and students
  should
• Be more interactive so that students can for
  example, make mutations in the structure
• Have better control on superposition of
  structures
• Have an undo command
• Be able to import and export more file formats
• Have both a menu driven and command line
  interface
• have different interfaces for research and
  education and perhaps have a tunable interface
• Be a multifunctional suite of programs that can
  all read the same or related formats

22
Summary how do we proceed from here?

• How do we ensure that large biological molecules
  like viruses and ribosomes are uniformly
  represented in the PDB file?
• How do we create a channel of communication
  between the user community and visualization
  software developers in order to develop better
  visualization resources?