AmberTools Bug Fixes and Updates

Note: The individual bug fixes for AmberTools 12 and 13 are described below in case you want to see what they do, or you don't want to use update_amber or patch_amber.py for some reason. For older versions of AmberTools, see the instructions associated with that particular version, although we strongly suggest upgrading to the latest version since they are more actively supported.

Most users are encouraged to use the update_amber (or patch_amber.py) script (which is in $AMBERHOME, and is automatically run during the configure step) to handle patching tasks. This script automatically downloads and applies patches on this site individually and keeps track of every patch that has been applied, making patching and trouble-shooting easier.

Instructions for using update_amber can be found in the AmberTools 13 manual in Chapter 1 Section 5. See Jason's wiki page for more detailed instructions, as well as a work-around in case you are behind a firewall that blocks patch_amber.py from contacting ambermd.org. Note, for update_amber, there is built-in Proxy support (see the AmberTools 13 manual for instructions on setting up your Proxy).

If you wish to apply these updates, run the command

$AMBERHOME/update_amber --update

Version 13

23	April,  2013

• update.1: Multiple Cpptraj fixes:
  1. Fix dimensions of VolMap grid when centering on a mask
  2. Fix handling of box coordinates when a DCD file has big-endian bit ordering
  3. Fix clustering of data sets
  4. Remove vestigial variable
  5. Update version

• update.2: Prevents MMPBSA.py from ignoring changes to _MMPBSA_info that it deems may be permitted. Also fixes the checkValidity action in ParmEd.

• update.3: Multiple fixes to cpptraj, ParmEd, and calcpka:
  1. (cpptraj) Expanded error messages for topologies broken by tleap
  2. (cpptraj) Enable noheader option for gnuplot
  3. (cpptraj) Recognize rms keyword in clustering
  4. (cpptraj) Fix possible segfault in hist when giving same data name twice
  5. (cpptraj) Improve identificatino of string data sets in standard data files
  6. (ParmEd) Add to the setMolecules option to fix prmtop files broken by tleap when non-adjacent molecules are cross-linked by bonds
  7. (ParmEd) Add the ability to read and write Amber NetCDF Restart files.
  8. (calcpka) Fix the horrible formatting of the populations.dat file

Version 12

12	April,  2012

• bugfix.1: Fixes Python mask parser (used by ParmEd, xParmEd, and MMPBSA.py) so it properly recognizes wildcards for multiple atom name/type selection. Also add a "deleteDihedral" function to ParmEd for topology file modification.

• bugfix.2: (1) Cpptraj: Fix reading of Charmm PSF files, fix frame and atom #s in 'closestout' output of closest command, update version string to 12.1. (2) Ptraj: Update version string to 12.0.

• bugfix.3: Fixes a regular expression in patch_amber.py so it will match file names and directories with a period and + in them.

• bugfix.4: Fixes a number of bugs, mostly improving error messages. (1) RATTLE errors are now fatal in NAB (2) Fix a name clash in rism1d that prevented successful compiling with GNU 4.7 compilers (3) Fix CHAMBER real()-type declarations to be consistent, avoiding errors on some platforms (notably Mac OS X 10.7 with Xcode 4.3.2). (4) Correct MMPBSA.py error reporting when necessary programs are not present or AMBERHOME is not set. Also fix PBSA/decomp/stability input file.

• bugfix.5: Makes cpptraj 'closest' action behavior match ptraj (solute is always kept intact).

• bugfix.6: Fixes a number of small bugs for rism1d that were primarily seen on 32-bit operating systems. Also improves configure argument suggestions on Mac OS X.

• bugfix.7: Fixes Intel 12 compiler problem for FFTW 3.3.

• bugfix.8: Fix unresolved symbols compiling cuda code on some versions of Ubuntu and possibly other Linux flavors. Also attempt to work around missing mpi.h when compiling the MPI version of cuda the code using system installed MPI instances.

• bugfix.9: Cpptraj: -Fix how number of outputted frames is determined -Adds 'nastruct' keywords 'hbcut' to set cutoff for determining H-bond cutoff in base pairing (Angstrom), 'origincut' to set the base reference axes cutoff for determining base pairing (Angstrom), and 'ref' to use a reference structure for base pair identification -Fix issue where large PDB files can overflow atom number columns

  MMPBSA.py: Adjust to cpptraj fix above -- this fixes MMPBSA calculations with only 1 frame in the input trajectory

  ante-MMPBSA.py: Allow users to specify a new PBRadii set that is appropriate to the solvent model they want to use

• bugfix.10: Cpptraj: Correct how cpptraj handles DCD box information so that imaging is performed correctly. Also solves a possible situation in which cpptraj hangs indefinitely when trying to image a DCD file with an orthorhombic box.

• bugfix.11: Same as bugfix.8 but includes the GNU compiler which was inadvertently missed in the first version of this bugfix.

• bugfix.12: PBSA: Fix a bug in water-related energy decomposition; Eliminate round-off errors in the level set function.

• bugfix.13: There's a chance that configure may not apply all bug fixes if authorized to do so. This bugfix addresses that.

• bugfix.14: Fixes the changeLJSingleType command in ParmEd (both parmed.py and xparmed.py)

• bugfix.15: Fixes DELTA G binding calculations when nmode entropy is included in cases where fewer frames are included in the normal mode calculations than the implicit solvent calculations.

• bugfix.16: Add new action to cpptraj, autoimage. Performs automatic imaging of a system with little to no additional input from user.

• bugfix.17: Fixed bug missing newline bug when writing DX format files. Fixes units of potential maps written in DX format. Will now be properly scaled to kcal/mol-e rather than the unscaled array values.

• bugfix.18: Fix potential segfault in autoimage when preceded by a strip command that removes all solvent. Also add the charmm keyword back to trajout.

• bugfix.19: Fix d orbital bug in semiempirical code: a) Use d orbitals with MNDO/d also for Br and I, b) Fix Slater-Condon parameter calculations for d orbitals with principal quantum number n>4

• bugfix.20: Remove pol12 force field files that are not yet ready for release.

• bugfix.21: Cpptraj: Fix bug where reading restart files with velocity info after an initial restart with no velocity info would result in corruption of box coordinates.

• bugfix.22: patch_amber.py: Fixes support for compressed patches necessary for the large Amber 12 CUDA patch.

• bugfix.23: configure, MMPBSA.py tests: Updates the configure script to work with the large CUDA upgrade in Amber 12 (bugfix.9). Also updates the MMPBSA.py tests to reflect the tweak to the igb=5 parameters in sander and pmemd.

• bugfix.24: MMPBSA.py: Recover error messages printed by PBSA to stdout. This doesn't fix PBSA errors, but it will now print error messages to the screen during MMPBSA.py calculations to aid in debugging.

• bugfix.25: MMPBSA.py: fix alanine scanning when mutant_only=1

• bugfix.26: Cpptraj: Fix reading of Charmm PSF files when atom type is not an integer. Also add detection of DCD # frames by file size to avoid spurious errors in the case where # of frames reported by DCD file does not match actual # of frames.

• bugfix.27: NAB and MMPBSA.py: Fix reporting of EDISPER in NAB programs (including mmpbsa_py_energy) when inp is 2.

• bugfix.28: Fix comments in leaprc.ff12SB to more correctly describe what files are used.

• bugfix.29: Fix bug with two-electron integrals in sqm for atoms separated by more than 10 Angstrom.

• bugfix.30: Fixes error with QM/MM switching function in presence of link atoms (affects AMBER sander simulations that use sqmlib)

• bugfix.31: Cpptraj: Fix infinite loops in orthorhombic dist calc/imaging when box lengths == 0.0 (bug 209); fix hbond autosearch when either only donor or only acceptor mask specified; fix parsing of input files with DOS-style lines; fix memory leak in DCD box read.

• bugfix.32: MMPBSA.py: Fixes defaults for APBS calculations, fixes calculations using sander that require inp=1, and improves error messages for incompatible topology files used with alanine scanning.

• bugfix.33: MMPBSA.py: Expand the list of allowable QM models to match SQM when running QM/MMGBSA calculations.

• bugfix.34: ParmEd (xparmed.py, parmed.py): Fix stripping atoms from the middle of a topology file when the coordinates from a restart file are present.

• bugfix.35: SQMLIB: QM/MM Ewald bug fixes (affects AMBER sander simulations that use sqmlib, see also AMBER bugfix.15; has no effect on sqm)
  1. add self-plasma term for charged systems
  2. improve default values for k space vectors

• bugfix.36: MMPBSA.py: Fix to allow chamber topology files to work with MMPBSA.py

• bugfix.37: patch_amber.py and update_amber: A re-write of the automatic updating/patch applying script with many performance and stability improvements.

Gaussian 09 fix

In Gaussian09 rev B.01, the facility to write out the electrostatic potential on a grid of points was inadvertently deleted. This means that antechamber and resp jobs won't work as they should. Fernando Clemente of Gaussian has kindly provided a script to work around the problem. Download the fixreadinesp.sh file, and follow the instructions there. (Note: you will have to make the script executable by typing chmod +x fixreadinesp.sh.)

Version 1.5

1.5	April,  2011 (git tag: at15)

• bugfix.all: This patch script contains all bug fixes for AmberTools version 1.5 in a single patch script. You can patch your AmberTools distribution to include all of the current bug fixes by downloading this file and saving it in your $AMBERHOME directory. You should then patch your system by running:

  cd $AMBERHOME
  patch -p0 -N < bugfix.all
  

  This should patch your source files with the latest bugfixes. Any bugfixes that appear to have already been applied will be skipped. Once the patching is finished you need to recompile your copy of AmberTools.

• bugfix.1: Enhancements and fixes for cpptraj. From version 1.0.5 to 1.0.8.

  Enhancements: (1) Added "average" keyword to "reference" command; allows one to average the coordinates of an input trajectory and use as a reference struct. (2) Added "precision" keyword to "datafile" command; users can specify an output width and precision for data sets in a data file. (3) Added "maxmin" keyword to "outtraj" command; users can direct outtraj to only print frames that are between a specified max and min based on specified data set.

  Bug fixes: (1) Box type detection is now fixed for general triclinic cells. (2) Fixed segfault when reading trajectory with box coords when associated parm has no box informaton. (3) Fix potential write of binary characters to Amber trajectory files when using "outtraj". (4) General improvement in output and debugging messages.

• bugfix.2: MMPBSA.py failed to mutate the residues ILE, THR, and VAL. Also, setup.sh in mmpbsa_py failed to set AMBERHOME to the correct value if it was not set at all. Fixes both issues
• bugfix.3: Adds variable 1-4 scaling to tleap
• bugfix.4: Cpptraj (V1.0.8 -> V1.0.9): Fix bug that could result in long wait times during file type detection.
• bugfix.5: Fix atom type definition for iodine in antechamber (Amber types)
• bugfix.6: AT15_Amber11.py: Inform users of tests that are expected to fail and/or cause an error with AmberTools 1.5 + Amber11.
• bugfix.7: readparm.py: add support for chamber-created prmtop files and fix LJ indexing issues for extracting original radii/well depths from the A/BCOEF arrays.
• bugfix.8: Support for FedoraCore 15 and other machines using gcc 4.6.
• bugfix.9: Ptraj: Fixes problem reading compressed Amber Trajectories with offset values > 1.
• bugfix.10: Cpptraj: Restart/mol2/pdb output frame numbers now match datafile frame numbers. Also fix read of broken Amber trajectories.
• bugfix.11: configure: Fix configuration for installation of native Windows version of pmemd.

• bugfix.12: Fix H1-CT-C -O and related angles to ensure wild cards don't contribute.
• bugfix.13: Patch to configure to work with Amber 11 GPU updates (bugfix 17 in Amber 11)
• bugfix.14: Fix reading of gzipped Amber trajectories in ptraj/cpptraj with uncompressed size > 4GB.
• bugfix.15: Fix segmentation faults of some pbsa tests when building with Intel compilers version 12.
• bugfix.16: Cpptraj: Fix potential segmentation fault during Per-residue RMSD calc ('rmsd perres').
• bugfix.17: MMPBSA.py: Fix bug that prevents alanine scanning from working properly with MM/3D-RISM.
• bugfix.18: Ptraj: Several bugfixes. (1) Fix bug in 'closest' where wrong distances calcd if no box info present. (2) Fix minor memory error in 'secstruct'. (3) Correct behavior of 'vector principal' when 'x|y|z' not specified. (4) Fix segfaults in 'watershell' command.
• bugfix.19: Fix case where even valid options for inp are flagged as illegal choices in MMPBSA.py
• bugfix.20: Don't mix the ildn and ff99SBnmr fixes, so that leaprc.ff99SBnmr matches the original paper.
• bugfix.21: Fixes excess chemical potential polar/apolar decomposition in 3D-RISM calculations and a few other minor 3D- and 1D-RISM bugs.
• bugfix.22: Fixes compilation for some compilers by appling missing 'save' attribute.
• bugfix.23: PBSA: A new box was not created for the second snapshot when imin = 5; Box dimensions are inputted only when a ligand is present, so in other cases, a new box is always created.
• bugfix.24: configure: Automagically apply AT15_Amber11.py when necessary (if Amber 11 exists alongside AmberTools 1.5). Can be disabled via -nopatch.
• bugfix.25: Miscellaneous fixes to MTK++ and MCPB.
• bugfix.26: Cpptraj => V1.1.2. Fix reading of restart files with # atoms > 5 digits.
• bugfix.27: MMPBSA.py: Fix parsing of Normal Mode entropy output
• bugfix.28: MMPBSA.py: Fix calculation of standard deviations for normal mode entropy calculations.

Version 1.4

1.4	April,  2010 (git tag: v_11)

• bugfix.all: This patch script contains all bug fixes for AmberTools version 1.4 in a single patch script. You can patch your AmberTools distribution to include all of the current bug fixes by downloading this file and saving it in your $AMBERHOME directory. You should then patch your system by running:

  cd $AMBERHOME
  patch -p0 -N < bugfix.all
  

  This should patch your source files with the latest bugfixes. Any bugfixes that appear to have already been applied will be skipped. Once the patching is finished you need to recompile your copy of AmberTools.

  In order to enable the DFTB tests for sqm (introduced with bugfix.8) you have to run:

  cd $AMBERHOME
  chmod ugo+x AmberTools/test/check_slko_files.x
  

  Remember that DFTB will only work if you install the DFTB integral (Slater-Koster) parameter files.

• bugfix.1: Update atom types in gaff.dat; remove some unused parameters.

• bugfix.2: Fix a bug that can cause segfaults in sleap from the addions command.

• bugfix.3: Update tleap and xleap to avoid some segfaults on Fedora Core 12/13 and possibly Ubuntu 10.04.

• bugfix.4: Fix to parallel ptraj when compressed trajectories are input.

• bugfix.5: Fix to 3D-RISM results for ionic solvents (like NaCl in water).

• bugfix.6: The gbsa=1 option gives bad results when mme2() is used.

• bugfix.7: Fix possible data overflow in pbsa.

• bugfix.8: Enable DFTB in sqm.

• bugfix.9: Fixes issues in ptraj. 1) ptraj could not properly process large compressed and uncompressed trajectory files. 2) the cluster command could trigger a segmentation fault.

• bugfix.10: Second derivative calculations in NAB would fail with 6-fold dihedrals.

• bugfix.11: Enable d orbitals in the DFTB code.

• bugfix.12: Fixes bug in ptraj where reading an input file with no EOL would cause a segmentation fault.

• bugfix.13: Adds missing parameters for disulfide bonds in the amoeba force field.

• bugfix.14: Fixes performance issues with sqm. Calculations will be 30% faster after application of this patch.

• bugfix.15: Fixes issues in ptraj. 1) output of cluster trajectory could cause a segfault when trajectory became too large. 2) Fixes minor bug in secstruct (DSSP) command where turn populations would be slightly off for residues lacking one or more backbone atoms (H, C, N, or O) such as PRO or terminal residues.

• bugfix.16: Fixes reading of large (>2GB) files in ptraj. Without this patch very large Amber Trajectories may become corrupted (Note: NetCDF trajectories do not have this issue).

• bugfix.17: Makes the interface of the addions command in sleap more similar to tleap's.

• bugfix.18: Makes corrections or improvements to commands in sleap: saveAmberParm, savePdb, combine, sequence, set UNIT box, solvatebox, and check.

• bugfix.19: Fixes issue where sleap didn't print negative dihedral force constants.

• bugfix.20: N-terminal proline charges in ff03 are incorrect. Also adds ACE N-terminal residue to all_aminont03.lib

Version 1.3

1.3	December,  2009 (CVS tag: at_1_3)

• bugfix.all: This patch script contains all bug fixes for AmberTools version 1.3 in a single patch script. You can patch your AmberTools distribution to include all of the current bug fixes by downloading this file and saving it in your $AMBERHOME directory. You should then patch your system by running:

  cd $AMBERHOME
  patch -p0 -N < bugfix.all
  

  This should patch your source files with the latest bugfixes. Any bugfixes that appear to have already been applied will be skipped. Once the patching is finished you need to recompile your copy of AmberTools.

• bugfix.1: Make an scf failure a fatal error for sqm, to minimze wasted computer time on a failed calculation.

• bugfix.2: Allow the pbsa program to handle halogens in the GAFF force field.

Version 1.2

1.2	July,  2008 (CVS tag: at_1_2)

• bugfix.all: This patch script contains all bug fixes for AmberTools version 1.2 in a single patch script. You can patch your AmberTools distribution to include all of the current bug fixes by downloading this file and saving it in your $AMBERHOME directory. You should then patch your system by running:

  cd $AMBERHOME
  patch -p0 -N < bugfix.all
  

  This should patch your source files with the latest bugfixes. Any bugfixes that appear to have already been applied will be skipped. Once the patching is finished you need to recompile your copy of AmberTools.

• bugfix.1: antechamber test charmm ala2 output update for phase changes.

• bugfix.2: NAB failed assertion for some dihedral phases, such as in ff99bsc0.

• bugfix.3: fixes to ptraj related to netcdf, remd and compressed files

• bugfix.4: fixes buffer overrun when writing mol2 files in LEaP

• bugfix.5: remove some duplicate parameters in gaff.dat

• bugfix.6: fixes issues with ptraj failing to identify some CHARMM PSF files.

• bugfix.7: fixes issues with ptraj processing of REMD, netcdf trajectories

• bugfix.8: more fixes related to bugfix.7

• bugfix.9: don't allow dihedral phases that are not zero or pi (as in parmbsc0)

• bugfix.10: patch the loadMol2 command in LEaP so that it won't choke on odd residue numbers

• bugfix.11: fix to am1-bcc bond orders (and hence charges) for SO2 and NO2 groups

• bugfix.12: update leaprc files to recognize atom and residue names in PDB 3 format

Version 1.0

1.0	April, 2008 (CVS tag: at_1_0)

• bugfix.all: This patch script contains all bug fixes for AmberTools version 1.0 in a single patch script. You can patch your AmberTools distribution to include all of the current bug fixes by downloading this file and saving it in your $AMBERHOME directory. You should then patch your system by running:

  cd $AMBERHOME
  patch -p0 -N < bugfix.all
  

  This should patch your source files with the latest bugfixes. Any bugfixes that appear to have already been applied will be skipped. Once the patching is finished you need to recompile your copy of AmberTools.

• bugfix.1: nblist.c in NAB won't compile with MPI turned on

• bugfix.2: The readparm() routine will throw an error when MPI is turned on

• bugfix.3: sleap will not read a leaprc file from an arbitrary path

• bugfix.4: sleap does not recognize the closeness parameter in solvateBox

• bugfix.5: sleap cleanups required for Mac OSX

• bugfix.6: sleap fixes to addIons to work with solvated systems

• bugfix.7: Fix to possible segfaults when doing clustering in ptraj