pdb_extract

 

pdb_extract - Workstation Version Manual

Extract information from X-ray crystallography, NMR, and EM software applications

(June, 18, 2004; last modified Nov 1, 2022) | (Latest version 4.0)
Table of Contents
What does pdb_extract do?     (TABLE OF CONTENTS)

pdb_extract is used to extract statistical information from the output files produced by software for protein structure determination using Xray Crystallography, NMR, and EM method. These statistical information will be written into a complete mmCIF file which is ready for PDB deposition.

In the case of Xray structure determination, pdb_extract merges all the information into two mmCIF (macromolecular Crystallographic Information File) files. One mmCIF file contains structure factors and the other contains atomic coordinates and statistics extracted from the steps of structure determination (data collection/integration/reduction, heavy atom phasing, molecular replacement, and final structure refinement) for various methods (MR, SAD, MAD, SIR, SIRAS, MIR, MIRAS). These two mmCIF files are ready for PDB deposition.

In the case of NMR and EM structure determination, experimental details from header section of PDB file and coordinates are extracted into a mmCIF file. This file along with chemical shift and restraint files are ready for PDB deposition.

The current version supports many software packages and hundreds of different output files produced in various of steps. Click here to see the supported software lists.

The assembled mmCIF files by pdb_extract should be used for Deposition.

The advantage of using pdb_extract:

  • Faster to prepare your mmCIF file for deposition. Users only provide the output files produced from various software to get all the statistics.
  • Complete and accurate to deposit your file. The statistics (ranging from index to final refinement) can be automatically extracted, so to avoid typos.
  • Great for multiple structural deposition by using the re-usable data template file
  • Both command options and Web interface are provided. It is flexible to use.
  • Collectively, these software tools reduce the human effort required to assemble complete and validated protein structure entries ready for PDB deposition.

IMPORTANT NOTES:

  1. The LOG or output files generated from any software should not be modified. Otherwise, information may not be extracted.
  2. If you have several structures ready to be deposited to the PDB site, you need to apply the pdb_extract program to each individual structure, since each structure requires a single PDB ID for deposition.
  3. You may have a lot of trials for each step (data processing, heavy atom phasing, or density modification, or final structure refinement), but information extracted from each step should be only from the best trial that leads to next step toward solving your structure.
  4. You may use different programs for heavy atom phasing solution. For example, you used program A to locate heavy atom positions and you used program B to refine heavy atom parameters (like x, y, z, occupancy and B factors etc.). Phasing statistics information will be extracted from the output of program B; therefore, pdb_extract should be applied to the output of program B.
  5. You may also use different programs for final structure refinement, but pdb_extract should be only applied to the program which leads to your final structure deposition.
Program access     TOP
The source code of pdb_extract can be downloaded from the address https://sw-tools.rcsb.org/apps/PDB_EXTRACT/source.html. The software is available under an Open Source license.

pdb_extract has been integrated into CCP4 and the CCP4i interface(Version 5.0 and above). Users can run pdb_extract under the CCP4 environment.

Installations     TOP

System Requirements:

  • platform Intel-Linux:
  • Python (>3.6)
  • C/C++ compilers
  • Installation of source code distribution     TOP
    
    Please refer to the README file in the package
    
    
    Run the program     TOP
    
    Please refer to the README file in the package
    
    
    Data collection/reduction     TOP

    This section is used to collect statistical information from the LOG files generated by the programs for Data Scaling/Merging/Averaging.

    Important: The log files must be generated from the LAST (or BEST) trial which corresponds to the files used for phasing or molecular replacement.

    The extracted information may be the following:

    
    *    Intensities (or amplitude)  and standard deviations 
    *    Data completeness (overall, resolution shells) 
    *    Redundancy (overall, resolution shells), mosaicity 
    *    R-merge, R-sym (overall, resolution shells) 
    *    average(I/sigma), (overall, resolution shells) 
    *    Total  and unique reflections collected. 
    *    Resolution range 
    
    

      Some helpful hints for getting LOG files from the program of Data Scaling/Merging/Averaging

    Using HKL/HKL2000/scalepack

    HKL (or HKL2000 or Scalepack) is a package by Otwinowski for data collection/reduction/scaling. You can use the graphical interface or the scalepack script to scale your data. The LOG file (e.g. scale1.log) contains statistics for PDB deposition.
    The generated LOG file type is 'LOG'.

    Using D*trek

    D*trek is a package by Jim Pflugrath at Rigaku/MSC for data collection/reduction/scaling. You can use the graphical interface to scale (or merge/average) your data. The LOG file (e.g. scale1.log) containing statistics is from the step of scaling data.
    The generated LOG file type is 'LOG'.

    Using SAINT

    SAINT is a package by Bruker (Siemens Molecular Analytical Research Tool) for data collection/reduction/scaling. The LOG file (e.g. scale1.ls) containing statistics is from the step of scaling data.
    The generated LOG file type is 'LOG'.

    Using SCALA

    SCALA/AIMLESS is the CCP4 supported program. It scales together multiple observations of reflections. SCALA generates mmCIF or LOG file containing useful statistics. When you run the programs, you must ask the program to export the data harvest file (mmCIF type). The mmCIF file will be name.scala or name.truncate. Otherwise, it will generate LOG file.
    The generated LOG file type is 'LOG or mmCIF'.

    Molecular replacement     TOP

    This section is used to collect key statistical information from Molecular Replacement. You may first generate a LOG file from the rotation function, then generate a LOG file from the translation function. You can upload the two LOG files into this section for data extraction. You can also upload one LOG file which is generated from MR.

    Important: The log files must be generated from the LAST (or BEST) trial which corresponds to the files used for density modification or refinement.


    The extracted information may be the following:
    *      Low and high resolution used in rotation and translation.
    *      Rotation and translation methods
    *      Reflection cut off criteria, reflection completeness.
    *      Correlation coefficients for I or F between observed and calculated.
    *      R_factor, packing information, and model details.
    

    Some helpful hints for getting LOG files from the program molecular replacement

    Using CNS/CNX/XPLOR

    CNS can be used to do molecular replacement. After you finish the translation search, you can get a log file called translation.list which contains all the information of molecular replacement.

    Using Amore (CCP4)

    Amore is a program for molecular replacement. It is distributed in the CCP4 package. After rotation and translation search, you will generate two log files rotation.log and translation.log. You may extract information from both log files

    If you run the program in one script, you may generate one LOG file. Upload this LOG file to the web interface.

    Using Molrep(CCP4)

    Molrep is a program for molecular replacement. It is distributed in the CCP4 package. When you run the script, you can specify a LOG file name (e.g. molrep.log). All the statistic information will be recorded in the log file.

    Using EPMR

    EPMR is a command line program for molecular replacement. When you run the program, please give a log file name like the following Epmr [options] files > epmr.log All the statisticial information will be written in the log file.

    Using Phaser

    Phaser was developed by Randy Read's group at the University of Cambridge. It is a program for phasing macromolecular crystal structures with maximum likelihood methods. The program generates a LOG file which can be uploaded to the web interface for data extraction.

    Heavy atom phasing     TOP

    Heavy atom phasing is performed at an earlier stage of structure determination. The log files generated from phasing contain important statistical information which should be deposited to the Protein Data Bank.

    From heavy atom phasing, you may have LOG files and heavy atom coordinate file.

    The phasing methods are the followings:
    *      MR   molecular replacement.
    *     SAD   single anomalous dispersion. 
    *     MAD   multiple anomalous dispersion.
    *     SIR   single isomorphous replacement.
    *   SIRAS   single isomorphous replacement with anomalous scattering.
    *     MIR   multiple isomorphous replacement.
    *   MIRAS   multiple isomorphous replacement with anomalous scattering.
    

    Important: The log files must be generated from the LAST (or BEST) trial which corresponds to the files used for density modification or refinement.


    The following items may be extracted:
    *     Wavelength, f_prime, f_double_prime, resolution range 
    *     FOM (acentric, centric, overall, resolution shells)
    *     R-Cullis (acentric, centric, overall, resolution shells)
    *     R-Kraut (acentric, centric, overall, resolution shells)
    *     Phasing power (acentric, centric, overall, resolution shells)
    *     Number of heavy atom sites, heavy atom type. 
    *     Heavy atom location method.
    *     Heavy atom B-factor, occupancies, and xyz coordinates.
    

    Some helpful hints for getting the output files generated by various programs

    Using SOLVE (version 2.00 and above):

    SOLVE is a program for finding heavy atom location and refining heavy atom parameters. The statistical information is written to a file solve.prt (default name used by the program). The heavy atom coordinates are written to a file ha.pdb.

    Note: You may upload the two file names solve.prt (file type: LOG) and ha.pdb (file type: PDB).

    Using CNS/CNX/XPLOR

    CNS is a complete software system for protein crystallography. The scripts for heavy atom location and phasing refinement are mad_phase.inp or ir_phase.inp. When you run these scripts, you will get output files like phase_final.summary, phase_final.sdb or mad_phase.fp.

    The output file phase_final.summary has all the phasing statistics.
    The output file phase_final.sdb has all the heavy atom coordinates, occupancies and B factors.
    The output file mad_phase.fp has refined f_prime and f_double_prime.

    (Note: The refined heavy atom coordinates, B factors and occupancies can be found in a file like phase_final.sdb. If you prefer to convert to the PDB format, you can run the script sdb_to_pdb.inp. You will get a file phase_final.pdb with PDB format.)

    Note: You may input at most three files (as shown above) for extracting phase information.

    Using MLPHARE (CCP4)

    MLPHARE is a program in the CCP4 suite. It is used for refining heavy atom parameters.

    If you use the CCP4i graphical interface or the script mode, you need to ask the program to write a harvesting file. Select the data havest button, when you use the CCP4i interface. Do not use the key word NOHARV, when you use script. After you finished running this program, you will get a file (e.g. name.mlphare) which is in mmCIF format. It contains all the information for heavy atom phasing refinement.

    For extracting the wavelength information, you need to run program REVISE in the CCP4 (version 4.0-4.2.2). You may get a file (e.g. prephadata.log)

    Note: You may input at most two files (as shown above) for extracting phase information.

    Using SHARP (version 1.3.x and 2.0 and above):

    SHARP is a program for finding heavy atom positions and refining heavy atom parameters. When you run SHARP or autoSHARP, the log files which have useful information are normally in the directory sharpfiles/logfiles_local/dirs, where dirs are all the subdirectories for your various structures. Please note that the location of generated log files may depend on how the program is installed!

    SHARP produces many output files.

    For version 1.3.x:
        Heavy.pdb  contains the heavy atom coordinates.
        FOMstats.html   contains figure of merit statistics.
        Otherstat.html  contains Rcullis, Rkraut, phasing power.
        
    For version 2.0 and above:
        Heavy.pdb   contains the heavy atom coordinates.
        FOMstats.html   contains figure of merit statistics.
        RCullis_?.html  contains Rcullis.
        PhasingPower_?.html  contains phasing power
    

    The easiest way to obtain these files is to run the program from the SUSHI interface. Review all the log files from the internet browser and save the files as plain text files.

    Note: You may input at most four files (as shown above) for extracting phase information.

    Using SnB (version 2.0 and above):

    SnB has no heavy atom parameter refinement, and it has no corresponding statistics. SnB gives the heavy atom or substructure coordinates (e.g. heavy.pdb) in PDB format.

    Note: You may input only one file (as shown above) for phasing extraction.

    Using BnP (version 0.93 and above):

    BnP is a combination of program SnB and Phases. The heavy atom positions are located by SnB and the heavy atom parameters will be refined by Phases.

    The log file (e.g. auto.log) can be found from the directory ~/PHASES/*. Log file normally contains phasing power for each phasing set.

    The file is in LOG format.

    Note: You may input at most one file (as shown above) for extracting phase information.

    Using SHELXD or SHELXS (version 97):

    Heavy atom or substructure coordinates are produced in PDB format (e.g. heavy.pdb).

    Note: You may input at most one file (as shown above) for extracting phase information.

    Density modification     TOP

    Density modification is normally performed after obtaining phases. If you do density modification in your structure determination, statistics information is needed for PDB deposition.

    If density modification is not done in a separate step, you may skip this step, since you do not have a log file specifically for density modification.

    Important: The log files must be generated from the LAST (or BEST) trial which corresponds to the file used for refinement.


        
    The following items may be extracted:
    *     Density modification method.
    *     FOM after density modification (overall, resolution shells)
    *     Solvent mask determination method.
    *     Structure solution software.
    

    Some helpful hints for getting the output files from each program:

    Using RESOLVE (version 2.00 and above):

    RESOLVE is a density modification program in the SOLVE/RESOLVE package. Normally it runs together with SOLVE, but one can run it separately. When you run RESOLVE, you will get a log file like resolve.log.

    Only one log file (resolve.log) is needed for extraction. File type is LOG.

    Using CNS/CNX/XPLOR

    The CNS user may need to run the input script like density_modify.inp. You will get a log file called density_modify.list.

    Only one log file (density_modify.list) is needed for extraction. File type is LOG.

    Using DM (CCP4)

    DM is a density modification program in the CCP4 suit. When you run DM either by using the CCP4i graphic interface or the script, you will get a log file like dm.log.

    Only one log file (dm.log) is needed for extraction. File type is LOG.

    Using SOLOMON (CCP4)

    SOLOMON is also a another density modification program in the CCP4 suite. When you run DM either by using the CCP4i graphic interface or the script, you will get a log file like Solomon.log.

    Only one log file (Solomon.log) is needed for extraction. File type is LOG.

    Final structure refinement     TOP

    Structure refinement is performed at the end of structure determination. The atom coordinates are generated in PDB or mmCIF format and the statistics are generated in log files. The pdb_extract program is applied to extract statistical information:

    Since statistics can be carried at the header section of PDB file, you may not provide any LOG files for some programs like CNS, REFMAC5.

    Important: The log file and the coordinate file must be generated from the LAST (or BEST) trial which corresponds to the file that is used for deposition to the PDB.


     
    The following items may be extracted:
    *    Resolution range (highest res. shell)
    *    Number of reflections used in refinement, and in R-Free set.
    *    R-factor (overall, resolution shells)
    *    Number of atoms refined
    *    Cell parameters and space group.
    *    The xyz coordinates of all the atoms.
    *    RMS Bond Distances, Bond Angles, Chiral Volume, Torsion Angles
    *    Isotropic temperature factor restraints
    *    Non-crystallographic symmetry restraints
    *    Solvent model used 
    *    Overall Average Isotropic B Factor
    *    Overall Anisotropic B Factor
    *    Overall Isotropic B Factor 
    *    Topology/parameter data used to refine deposited model
    *    Refinement software
    

    Some helpful hints for getting the output files from each program:

    Using REFMAC5 (CCP4):

    REFMAC5 is a program for structure refinement used in the CCP4 suite. If you run this program using CCP4i or the script, you can get a PDB file with all the refinement information at the header section.

    You may directly deposit this PDB file.

    Using CNS/CNX/XPLOR

    CNS/CNX/XPLOR is a program for final structure refinement. It exports coordinate file in both PDB and mmCIF format. You need the script deposit_mmcif.inp to generate the mmCIF format.

    The mmCIF file carries more statistical information than the PDB file. Authors are encouraged to deposit the mmCIF file, otherwise authors may need to manually fill in more information.

    You may not have to give any LOG file generated from CNS/CNX/XPLOR.

    Using SHELXL (version 97):

    SHELXL is a sub_program in the SHELX package. It is used for structure refinement. After you finish structure refinement, you need to run the shelxpro interactive program and use option B. After going through the shelxpro, you will get a PDB file (e.g. name.pdb) with header information.

    Using TNT (version 5f):

    TNT is a crystal structure refinement program. Data from this program can be extracted from the output PDB file and some LOG files. You can use the to_pdb command to convert coordinates in TNT format (name.cor) to the PDB format (name.pdb).

    The command is: to_pdb name.cor

    After finishing refinement, you must use command rfactor to generate a log file (e.g. rfactor.log) which contains the refinement statistics.

    The command is: rfactor name.cor > rfactor.log

    To extract the symmetry information, user must provide the symmetry file (e.g. p6122.dat). This information is in the control file name.tnt

    Using ARP/wARP:

    ARP/wARP is a automatic program for model building and refinement. REFMAC5 is used for the structure refinement step.

    The new version (6.0 or above) can use CCP4i as graphic interface. You can run this program either by CCP4i or by using script. You will get a log file (for example warpNtrace_refine.log). You also get a PDB file like warpNtrace.pdb.

    Note: If the coordinate file warpNtrace.pdb is directly used for deposition, you can use this option. Otherwise, use other program for final refinement.

    Using PHENIX

    PHENIX is a new software suite for the automated determination of macromolecular structures using X-ray crystallography and other methods.

    The PDB file generated by phenix.refine has the non-standard 'REMARK' and the standard 'REMARK 3'. It is also OK to keep the non-standard REMARK for deposion.

    Note: Sometimes, the MTZ file from PHENIX only contains 2Fo-Fc. Before deposition, you must make sure that the amplitude (Fo) or Intensity (I) is included in the MTZ file.

    Tables     TOP
    Below are the two Tables. One is for all the command options and the other is for the software supported by pdb_extract.
        TOP command options
    command line options consist of three executable components of pdb_extract.
    pdb_extract is used to capture the details of molecular replacement, heavy atom phasing, density modification and structure refinement.
    pdb_extract_sf is used to convert all other structure factor format to mmCIF format for PDB deposition,
    extract is used to generate a data template file (data_template.text) and a script file (log_script.inp).
    pdb_extract [OPTION]... [FILE]...
    Option
    Arguments followed by each option
    -o output file name (default name is pdb_extract.mmcif)
    -e one of experimental methods
    (MR| SAD | MAD | SIR | MIR | SIRAS | MIRAS)
    -i one of programs for indexing
    [HKL | DENZO | DTREK | MOSFLM]
    -s one of programs for reflection data scaling (used for refinement)
    [SCALA |AIMLESS HKL | SCALEPACK | DTREK | SAINT | 3DSCALE | XSCALE | XENGEN | PROSCALE]
    -m one of programs for molecular replacement
    [AMORE | CNS | XPLOR | EPMR | MOLREP | BEAST | PHASER | COMO]
    -p one of programs for heavy atom phasing
    [CNS | XPLOR | MLPHARE | SOLVE | SHELX | SNB | BnP | BP3 | SHARP | PHASER | PHASES | WARP]
    -r one of programs for final structure refinement
    [CNS | XPLOR | REFMAC5 | SHELX | TNT | BUSTER | PROLSQ | NUCLSQ | RESTRAIN | PHENIX | MAIN]
    -ipdb the input file with PDB format.
    -icif the input file with mmCIF format.
    -ient the input file data_template.text. (for complete sequence.)
    (It is generated by 'extract -pdb pdbfile')
    -idat the reflection data file to get < I/SigmaI > (optional)
    pdb_extract_sf [OPTION]... [FILE]...
    -o output file name (default name is pdb_extract_sf.mmcif).
    -rt data type (I or F) in the reflection data file
    (used for final structure refinement!)
    -rp One of data formats
    (CNS | mmCIF | SHELX | TNT | HKL/Scalepack | DTrek | SAINT | OTHER )
    (used for final structure refinement!)
    -dt data type (I or F) after data reduction at beam line.
    (used for phase determination!)
    -dp One of programs for data reduction ( HKL/Scalepack, DTrek, SAINT | OTHER ).
    (used for phase determination!)
    -c crystal number (like 1, 2, 3 ...) used for diffraction.
    -w wavelength number (like 1, 2, 3 ...) used for diffraction.
    -idat data file name used for phasing or structure refinement.
    Supported crystallographic software lists     TOP
    • PHENIX, REFMAC, CNS/CNX/XPLOR, BUSTER
    • AIMLESS, HKL-2000, HKL-3000, XSCALE, XDS, SCALA, SCALEPACK, XIA2, DIALS, POINTLESS, d*TREK