Pdb2pqr

Additions

* Created Sphinx documentation of usage and API at http://pdb2pqr.readthedocs.io (#88).
* New command line tools added with documentation (163).
* Added support for reading QCD-format structure files (137).
* Added versioneer support for versioning (104).
* Made several APBS tools available as PDB2PQR scripts: `dx2cube` (98), `inputgen` (105), `psize` (106).
* Added code of conduct document (62).

Fixes

* Fixed faulty no-op logic in debumping routines (162).
* Fixed problem with element type in PDB output (159).
* Updated very out-of-date change log (153).
* Fixed atom-ordering problem in PDB output (134).
* Fixed REDVAT PDB record parsing (119).
* Fixed broken `--apbs-input` option (94).
* Fixed OS-specific file handing (78).

Changes

* PDB2PKA is still removed from the code base while refactoring for a code base that is more friendly to multiple platforms.
* Added Python 3.9 to testing (161).
* Enabled additional PROPKA output (143).
* Moved mmCIF support to external module `mmcif-pdbx` (135).
* Added formal PQR parser (97).
* Made failure due to missing backbone atoms more graceful (95).
* Moved some logging output from stdout/stderr to files (74).
* Increased testing (70, 73).
* Continued de-linting and refactoring (56, 122).

Additions

* Added ability to read mmCIF files.

Fixes

* Updated URL used to fetch PDB files from RCSB.
* Fixed naming error for CYS hydrogen.
* Replaced Python pickle with portable JSON.
* Fixed packaging problem

Changes

* Upgraded to Python 3.
* Upgraded web interface.
* Upgraded to PROPKA 3.1 (and converted to :mod:`pip` dependency rather than submodule).
* Removed PDB2PKA support.
* Removed support for extensions.
* Significant code refactoring.
* Changed output from `print` to `logging`.
* Provided additional warnings when dropping HETATM entries.
* Improved build system.
* Increased list of proteins used in testing.
* Removed Opal support.
* Added GitHub actions for continuous integration testing.

These are notes for the current version of PDB2PQR

Please see http://www.poissonboltzmann.org/pdb2pqr/release-history for the complete release history

NEW FEATURES
- Replaced the Monte Carlo method for generating titration curves with graph cut. See http://arxiv.org/abs/1507.07021

BUG FIXES
- Added a check before calculating pKas for large interactions energies.

CHANGES
- The networkx library is now required for pdb2pka.

KNOWN BUGS
- If more than one extension is run from the command line and one of the extensions modifies the protein data structure it could affect the output of the other extension. The only included extensions that exhibit this problem are resinter and newresinter.
- Running ligands and PDB2PKA at the same time is not currently supported.
- PDB2PKA currently leaks memory slowly. Small jobs will use about twice the normally required RAM (ie ~14 titratable residues will use 140MB). Big jobs will use about 5 times the normally required RAM ( 60 titratable residues will use 480MB ). We are working to fix this.

Please see http://www.poissonboltzmann.org/pdb2pqr/release-history for the complete release history

Notable new features:

PDB2PKA as an alternative to PROPKA for calculating pH values used to protonate residues. This feature is EXPERIMENTAL. The libraries to make this feature available are included in the binary releases. They are NOT included in the source code and are not compiled with the rest of PDB2PQR.

Improved web interface.

NEW FEATURES
- Improved look of web interface
- Option to automatically drop water from pdb file before processing.
- Integration of PDB2PKA into PDB2PQR as an alternative to PROPKA.
- Support for compiling with VS2008 in Windows.
- Option to build with debug headers.
- PDB2PKA now detects and reports non Henderson-Hasselbalch behavior.
- PDB2PKA can be instructed whether or not to start from scratch with --pdb2pka-resume
- Can now specify output directory for PDB2PKA.
- Improved error regarding backbone in some cases.
- Changed time format on querystatus page
- Improved error catching on web interface.

BUG FIXES
- Fixed executable name when creating binaries for Unix based operating systems.
- Fixed potential crash when using --clean with extensions.
- Fixed MAXATOMS display on server home page.
- PDB2PKA now mostly respects the --verbose setting.
- Fixed how hydrogens are added by PDB2PKA for state changes in some cases.
- Fixed psize error check.
- Will now build properly without ligand support if numpy is not installed.
- Removed old automake build files from all tests ported to scons.
- Fixed broken opal backend.

CHANGES
- Command line interface to PROPKA changed to accommodate PDB2PKA. PROPKA is now used with --ph-calc-method=propka. --with-ph now defaults to 7.0 and is only required if a different pH value is required.
- --ph-calc-method to select optional method to calculate pH values used to protonate titratable residues. Possible options are "propka" and "pdb2pka".
- Dropped support for compilation with mingw. Building on Windows now requires VS 2008 installed in the default location.
- Updated included Scons to 2.3.3
- PDB2PKA can now be run directly (not integrated in PDB2PQR) with pka.py. Arguments are `PDB file` and `Output directory`.
- No longer providing 32-bit binary builds. PDB2PKA support is too memory intensive to make this practical in many cases.

KNOWN BUGS
- If more than one extension is run from the command line and one of the extensions modifies the protein data structure it could affect the output of the other extension. The only included extensions that exhibit this problem are resinter and newresinter.
- Running ligands and PDB2PKA at the same time is not currently supported.
- PDB2PKA currently leaks memory slowly. Small jobs will use about twice the normally required RAM (ie ~14 titratable residues will use 140MB). Big jobs will use about 5 times the normally required RAM ( 60 titratable residues will use 480MB ). We are working to fix this.

Please see http://www.poissonboltzmann.org/pdb2pqr/release-history for the complete release history

Notable new features:

PDB2PKA as an alternative to PROPKA for calculating pH values used to protonate residues. This feature is EXPERIMENTAL. The libraries to make this feature available are included in the binary releases. They are NOT included in the source code and are not compiled with the rest of PDB2PQR.

Improved web interface.

NEW FEATURES
- Improved look of web interface
- Option to automatically drop water from pdb file before processing.
- Integration of PDB2PKA into PDB2PQR as an alternative to PROPKA.
- Support for compiling with VS2008 in Windows.
- Option to build with debug headers.
- PDB2PKA now detects and reports non Henderson-Hasselbalch behavior.
- PDB2PKA can be instructed whether or not to start from scratch with --pdb2pka-resume
- Can now specify output directory for PDB2PKA.
- Improved error regarding backbone in some cases.
- Changed time format on querystatus page
- Improved error catching on web interface.

BUG FIXES
- Fixed executable name when creating binaries for Unix based operating systems.
- Fixed potential crash when using --clean with extensions.
- Fixed MAXATOMS display on server home page.
- PDB2PKA now mostly respects the --verbose setting.
- Fixed how hydrogens are added by PDB2PKA for state changes in some cases.
- Fixed psize error check.
- Will now build properly without ligand support if numpy is not installed.
- Removed old automake build files from all tests ported to scons.
- Fixed broken opal backend.

CHANGES
- Command line interface to PROPKA changed to accommodate PDB2PKA. PROPKA is now used with --ph-calc-method=propka. --with-ph now defaults to 7.0 and is only required if a different pH value is required.
- --ph-calc-method to select optional method to calculate pH values used to protonate titratable residues. Possible options are "propka" and "pdb2pka".
- Dropped support for compilation with mingw. Building on Windows now requires VS 2008 installed in the default location.
- Updated included Scons to 2.3.3
- PDB2PKA can now be run directly (not integrated in PDB2PQR) with pka.py. Arguments are `PDB file` and `Output directory`.
- No longer providing 32-bit binary builds. PDB2PKA support is too memory intensive to make this practical in many cases.

KNOWN BUGS
- If more than one extension is run from the command line and one of the extensions modifies the protein data structure it could affect the output of the other extension. The only included extensions that exhibit this problem are resinter and newresinter.
- Running ligands and PDB2PKA at the same time is not currently supported.
- PDB2PKA currently leaks memory slowly. Small jobs will use about twice the normally required RAM (ie ~14 titratable residues will use 140MB). Big jobs will use about 5 times the normally required RAM ( 60 titratable residues will use 480MB ). We are working to fix this.

Notable new features:

Binary builds.

Binary builds do not require python or numpy be installed to use. Everything needed to run PDB2PQR is included. Just unpack and use.

OSX binaries require OSX 10.6 or newer. The OSX binary is 64-bit.

Linux binaries require CentOS 6 or newer and have been tested on Ubuntu 12.04 LTS and Linux Mint 13. If you are running 64-bit Linux use the 64-bit libraries. In some cases the needed 32-bit system libraries will not be installed on a 64-bit system.

Windows binaries are 32 bit and were built and tested on Windows 7 64-bit but should work on Windows XP, Vista, and 8 both 32 and 64-bit systems.

Windows support

PDB2PQR can now be compiled and run on Windows using MinGW32. See http://mingw.org/ for details.

Compilation with Scons

PDB2PQR now uses Scons for compilations. With this comes improved automated testing.

Please see http://www.poissonboltzmann.org/pdb2pqr/release-history for the complete release history

NEW FEATURES
- A ligand file with duplicate atoms will cause pdb2pqr to stop instead of issue a warning. Trust us, this is a feature, not a bug!
- Improved error reporting.
- Added support for reference command line option for PROPKA.
- Added newresinter plugin to provide alternate methods for calculating interaction energies between residues.
- Mol2 file handling is now case insensitive with atom names.
- PROPKA with a pH of 7 is now specified by default on the web service.
- Compilation is now done with scons.
- Verbose output now includes information on all patches applied during a run.
- Added stderr and stdout to web error page.
- Added warning to water optimization when other water is ignored.
- Command line used to generate a pqr is now duplicated in the comments of the output.
- Added support for NUMMDL in parser.
- Added complete commandline feature test. Use complete-test target.
- Added propka support for phosphorous sp3. - Thanks to Dr. Stefan Henrich
- Added a PyInstaller spec file. Standalone pdb2pqr builds are now possible.

BUG FIXES
- Rolled back change that prevented plugins from interfering with each other. Large proteins would cause a stack overflow when trying to do a deep copy.
- Updated INSTALL file to reflect no more need for Fortran.
- Fixed apbs input file to match what web interface produces.
- Fixed user specified mobile ion species not being passed to apbs input file.
- Removed ambiguous A, ADE, C, CYT, G, GUA, T, THY, U, URA as possible residue names.
- Removed eval from pdb parsing routines.
- Updated web links to refer to http://www.poissonboltzmann.org where appropriate.
- Fixed hbond extension output to include insertion code in residue name.
- Fixed debumping routines not including water in their checks. Fixes bad debump of ASN B 20 in 1gm9 when run with pH 7.0.
- Fixed debumping failing to use best angle for a specific dihedral angle when no tested angles are without conflict.
- Fixed debumping using asymmetrical cutoffs and too large cutoffs in many checks involving hydrogen.
- Fixed debumping accumulating rounding error while checking angles.
- Fixed inconsistencies in pdb parsing. - Thanks to Dr. Stefan Henrich
- Fixed problems with propka handling of aromatic carbon/nitrogen. - Thanks to Dr. Stefan Henrich
- Fixed case where certain apbs compile options would break web visualization.
- Fixed improper handling of paths with a '.' or filenames with more than one '.' in them.

CHANGES
- Removed numpy from contrib. The user is expected to have numpy installed and available to python at configuration.
- Support for numeric dropped.

KNOWN BUGS
- If more than one extension is run from the command line and one of the extensions modifies the protein data structure it could affect the output of the other extension. The only included extensions that exhibit this problem are resinter and newresinter.