AMBER 8 Bug Fixes
8.0 March, 2004 (CVS tag: v8_0)
This patch script contains all of the bug fixes detailed below in a single patch script. You can patch your amber 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:
patch -p0 -N -r patch_rejects <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 amber.
LEaP may crash when writing a prmtop for a unit with unbonded hydrogens.
The configure option -scali does not produce MPI executables.
Antechamber will mis-read pdb files that have a non-blank chainID.
PBSA can fail with a "PB Bomb in epsbnd()" error message.
The configure targets hpf90 and HP-UX do not produce a functional xleap.
antechamber will generally fail when the input is a mol2 file; plus
other smaller antechamber fixups.
Using a restraintmask or a bellymask that invokes distance comparisons
A follow-on to bugfix.6: needed if you use antechamber with both the
-fi mol2 and the -j 5 flags at the same time.
The ioutfm = 1 option in sander is broken.
The leaprc.ff03 file loads old versions of ACE and NME protecting
New format prmtop files with more than 999999 lines are read incorrectly.
Having zero masses for extra points can cause problems on some machines
Several fixes to sander.QMMM
LEaP can print out incorrect warnings about perturbed charges
Improved error reporting in transforming chirality orientations in LEaP.
Parallel sander might fail on some platforms due to a bad broadcast
Ptraj may give spurious errors about not being able to open a file for
output when writing amber restrt files.
Group specifications don't work correctly when igrp(i) < 0.
Charges for neutral lysine (LYN) are incorrect in the ff03 force field.
The glycam04EP.dat file has some spurious lines that need to be removed.
Remove empty Fortran statements that NEC compilers (and others?) won't
The anal program prints an incorrect error message for
Constant pH simulations with only a single titrating site may fail.
Use of the grid command in ptraj will sometimes fail.
Langevin dynamics will fail when the target temperature is changed during
Sander.QMMM MD simulations involving link atoms may fail.
Sander fails to ignore the dielc variable for GB calculations;
this can lead to problems with mm_pbsa if the interior dielectric is not
The gaff.dat parameter file has some duplicate and erroneous entries.
Fixes possible segmentation fault in ambpdb. Also adds dynamic memory allocation to ambpdb.
If the Intel C/C++ compilers are used, icpc should be used for
the bondtype program, not icc.
The pmemd program can write unnecessary information to
fort.35 if DUMPAVE is requested.
The mden file can contain a bad value for pressure when the
pressure is not being regulated.
The forces for "cap" water are not accompanied by a corresponding energy
term. (See also bugfixes 40, 43 and 45, below.)
The OMIT_NAMES option to prepgen may not work correctly.
The pmemd program won't compile with the Portland group compiler.
The GROUP commands don't work in nmode.
Some targeted MD tests are not suitable for running in parallel and may make the test procedure fail.
Modifications to pmemd for the pathscale compilers.
The configure targets sparc and SunOS do not produce a functional xleap.
Same as bugfix.33, but also for sander.QMMM
Resetting an atom's pertType in tleap does not update the PERTURBED flag.
Modify leap's bond command to use S,D,T,A for single, double, triple and aromatic bonds
respectively instead of -,=,#,:. The old method might not work on all machines.
Updates the test case results that were changed as a result of bugfix.33
LEaP may crash when loading a prep file with LOOP bonds between
atoms of type dummy.
Updates another test case result that was changed as a result of bugfix.33
Protects pmemd from very rare error involving incorrect fractional
Fixes MPI calls in sander when the noBTREE is set during
Fix potential error with nmr weight restraints where value1 or value2 = 0.0.
The pbsa program can output highly fluctuating reaction field
energies for neutral systems.
The sander program can produce erroneous errors related to FFT dimensions
for large non-periodic systems.
Use of TIP3PBOX or SPCBOX directly (not via the solvateoct or solvatebox commands) in Leap can create a inpcrd file with incorrect box angles.
Manual setting of lastrst needs to be allowed for regular Ewald
Leap: Improper torsions specified in a frcmod file may not trump the same
improper torsions defined in the main parmxx.dat file.
The ntt=2 option in sander can result in average temperatures
that are too high.
The pmemd doesn't properly echo the title of new-style prmtop
Update to bugfix.54, which failed to work properly in parallel.
Equivalent to bugfix.11, for pmemd.
Equivalent to bugfixes 32, 54, and 56, for pmemd.
Sander with MMTSB support may fail to compile.
Provide 64-bit support for xleap
Solvent accessibility calculations with pbsa may fail on some
The ambpdb program doesn't allocate enough space for an internal
The iwrap=1 option in pmemd does not work correctly.
A workaround to ifort compiler bugs on itanium (ia64) machines