AMBER Page
AMBER 4.1 Bug Fixes
4.1 March 31, '95 (RCS 6.0)
- bugfix.1:
sander: ewald - pressure scaling & mdcrd box
- bugfix.2:
sander: Unicos 8.0 vector compilation ** modified
- bugfix.3:
gibbs: Unicos shared memory parallel resnba index
- bugfix.4:
1994 force field: reorder atoms in one improper torsion ** modified
- bugfix.5:
spasms: Makefile rule prevents building
- bugfix.6:
sander: extra open of unit 7 crashes on rs6000
- bugfix.7:
makeCHIR_RST: corrected /dev/tty path, better awk selection
** modified
- bugfix.8:
sander: MPI parallel problem starting MD w/out velocities
- bugfix.9:
spasms test scripts
- bugfix.10:
xleap: crashes on use of 'About.. more' button
- bugfix.11:
resp: matrix singularity err (fitting over multiple molecules)
** modified
- bugfix.12:
dshift: rprint option always on
- bugfix.13:
gibbs: no 1-4 polar energies in perturbed group (ipol=1)
- bugfix.14:
leap: errors reading user-supplied parm.dat files
- bugfix.15:
leap: mixed dihedral terms when using multiple parmsets
- bugfix.16:
leap: ions have unknown position in addIons
- bugfix.17:
makeNOEEXP: output format is wrong
- bugfix.18:
prep, parm: 'central atom 3rd' rule for improper torsions
- bugfix.19:
sander: turn off sgi_mp shake
- bugfix.20:
nucgen: use 1994 ff atom names for phosphate & sugar oxygens
- bugfix.21:
all_nuc94.in: wrong default placement of RNA H3'
- bugfix.22:
sander: MD Nonsetup CPU time only reported for last run, NRUN>1
- bugfix.23:
carnal: carbons can be identified as hbond donors w/ '94 ff
- bugfix.24:
sander: SP2/MPI compilation & correct i/o & dynamics
- bugfix.25:
carnal: PUCKER NUCLEIC fails to recognize '94 ff residues
- bugfix.26:
carnal: TORSION BACKBONE crash with small molecule at end
- bugfix.27:
carnal: appending to mdcrd output when -O used
- bugfix.28:
carnal: reading wrong numbers in cases like "160.3661090.239"
- bugfix.29:
carnal: appending to TABLE output when -O used
- bugfix.30:
all_aminont94.in: HB2,HB3 on PHE, TYR (symptom: missing CT-HA params)
- bugfix.31:
leap: slices of vacuum appear when solvating with WATBOX216
- bugfix.32:
sander: temperature blowup on SGI parallel
- bugfix.33:
sander, gibbs, nmode, anal: fractional dielectric silently ignored
- bugfix.34:
'94 ff: Missing parameters and wrong alpha-hydrogen atom type in neutral
ASP (ASH) and GLU (GLH)
- bugfix.35:
gibbs: -O option doesn't work with PATNRG file
- bugfix.36:
sander: wrong result / crash with NTT<0 (velocity reassignment)
- bugfix.37:
anal: fixes bugfix.33, which accidentally turned off all dielectric scaling
- bugfix.38:
leap: saveamberparmpert fails when no atoms are perturbed
- bugfix.39:
carnal [superseded by #56]
- bugfix.40:
carnal [superseded by #56]
- bugfix.41:
sander: NMR chemical shift options - ring current component
- bugfix.42:
gibbs: polar option fails with improper containing H
- bugfix.43:
edit: GEN option doesn't generate correct generic solvent box
- bugfix.44:
xleap: crashes when saving unit atom params (some sites)
- bugfix.45:
protonate: crashes on 'computed goto' error
- bugfix.46:
leap: crashes on superfluous charges in prep.in CHARGE section
- bugfix.47:
all_aminont94.in, parm94.dat: N-terminal PRO missing parameters
- bugfix.48:
sander: relative weight changes to VDW cause crash or wrong result
- bugfix.49:
leap: IDIVF field in Improper Torsions not ignored (should be)
- bugfix.50:
anal: added array bounds check for simulations w/ added params
- bugfix.51:
sander: with NTT<0, velocities set to same value on each reset
- bugfix.52:
carnal: user failure to complete COORD definition can cause crash
- bugfix.53:
sander: wrong coords for box in mden file
- bugfix.54:
all_amino*94.in: order of some impropers changed for leap
- bugfix.55:
anal: torsion energies can be slightly negative (e.g. -1e-9)
- bugfix.56:
carnal: reading pdb's w/ long/short lines & skipping non-atom lines
- bugfix.57:
carnal: per-atom & per-residue RMS wrong/crashes, tables with many items
- bugfix.58:
intense: crashes, missing option description
- bugfix.59:
carnal: sqrt() domain errors in RMS summary output on very small #s
- bugfix.60:
gibbs: MPI/MPICH parallel fixes
- bugfix.61:
sander & gibbs: unrequested removal of COM movement when applying
random velocities
- bugfix.62:
carnal: EXH2O option of COORD
- bugfix.63:
carnal: hangs reading a PDB file when prmtop has a box
- bugfix.64:
gibbs: overflow of step count in restrt file
- bugfix.65:
gibbs: possible overflow of unused vars in restrt file
- bugfix.66:
gibbs: MPI parallel allows running w/ unimplemented polar option
- bugfix.67:
carnal: using explicit prmtop with STATIC; printing prm title
- bugfix.68:
parm: use of frcmod for dihedrals when setting up perturbation
- bugfix.69:
gibbs: pert torsion wrong w/ Thermodynamic Integration
- bugfix.70:
sander/gibbs: IBM SP2 compilation
- bugfix.71:
carnal: use of group%{component} for point measurement
- bugfix.72:
prep: prevent use of '-M' for 3rd or 4th atom in IMPROPER section
- bugfix.73:
all_*94.in: default H atom positions made to agree w/ PDB convention
- bugfix.74:
all_nuc94.in: cytosine default H2' atom positions agreed w/ PDB convention
- bugfix.75:
gibbs: report of free E components/derivs wrong when dual cutoff
- bugfix.76:
all_aminont94.in: old ff charges on N-terminal THR, ARG, HIP, LYS.
- bugfix.77:
sander: center of mass motion removal when periodic boundary
- bugfix.78:
sander: update box properly for Ewald (SHAKE errors)
- bugfix.79:
sander, gibbs: non-TIP3 water w/ belly may use TIP3 bonds
- bugfix.80:
gibbs: wrong energies for FEP when NCORC=1 plus other special conds
- bugfix.81:
sander: crash when #/atoms small in relation to residue size
- bugfix.82:
carnal: RMS cmd references to individual sets in STATIC fail
- bugfix.83:
carnal: STATIC: problems using trajectory w/ box
- bugfix.84:
carnal: TORSION BACKBONE: problems with multiple molecules
- bugfix.85:
gibbs: (MPI parallel) crash when doing PMF
- bugfix.86:
gibbs: (MPI parallel) crash, erroneous free energies
- bugfix.87:
gibbs: (MPI parallel) crash or wrong results on bond calcs
- bugfix.88:
sander: (SGI parallel) wrong energies w/ dual cutoff on >1 cpu
- bugfix.89:
carnal: STATIC: can't reference individual sets after 1st
- bugfix.90:
carnal: HBOND DONOR group: can crash depending on group
- bugfix.91:
data_base: all_amino nt94.dat charge not integer value
- bugfix.92:
gibbs: Cap water indicies fixed
Untested gibbs/mpi bugfixes - these have been promoted to bugfixes 85-87.
Untested sander/mpi bugfix
Sander upgrade (6/19/95) - users with a previous 4.1 may request
an upgrade package for Sander. Two features are included.
Dynamic memory allocation causes the code to auto-dimension for
problem size and saves manual redimensioning.
An Ewald speedup for SGI R8000 machines on one user's
14000-atom system on 1 CPU reduced runtime from 680->470 seconds.
Send mail to amber-request@cgl.ucsf.edu to request the upgrade.
Porting changes (i.e. fixes to problems not counted as bugs
because machine-specific)