|
p. 7 |
In step (1) of the installation instructions: you need to (not just "probably want to") place $AMBERHOME/exe into your PATH variable. |
|
p. 13 |
For the bpti example: (1) you need to also remove alternate configurations in residue 39 as well as residue 50; (2) you should remove all of the CONECT records, since some of them now refer to atoms that have been removed, and LEaP will get confused; (3) finally, in the sample input, you will need to add source leaprc.ff94 (or similar) before the first command, since there is no longer a default leaprc file. |
|
p. 60 |
The references to gblambda under PBradii are incorrect. We recommend using bondi when igb=2, and mbondi when igb=1. |
|
p. 78 |
The second resp example has the wrong input and
output files; it should read:
|
|
p. 80 |
After the polarization equation, it should say that mu is an induced dipole (not an atomic polarizability). |
|
p. 102 |
Polarizable potentials do work with PME; the text should read: polarizability can be run with ew_type = 0, 1, or 2. |
|
p. 139 |
The variables fj and dfac_al in the example are obsolete; in Amber 7, you would use gigj instead. |
|
p. 167 |
The units of TAUP are ps (not ps**-1). |
|
p. 8 |
Step (4) is no longer necessary; roar-2.1 will be installed with the "make install" script from step 3. |
|
p. 11 |
For installing on Windows (bottom of the page), add the following: you should edit the amber7/src/sfx.h file to set $SFX=.exe, and use Machine.g77 as your MACHINE file. You will need to install the tcsh package from cygwin, and copy tcsh.exe to /bin/csh.exe. |
|
p. 12 |
Note that the atomic charges in the antechamber tests depend on mopac, and that different versions of this program (or even the same version on different machines) often give slightly different answers. You should expect to see slightly different charges (`+-` 0.01) on machines that differ from those on which we ran the tests (SGI R10000, using mopac509mn). |
|
p. 75 |
Add a fifth note at the bottom of the page: (5) The atoms in the input pdb file are all treated as being in a single residue, even if there are different residue names present in the pdb file. Further, the atom names should be unique, so that no two atoms have the same name. |
|
p. 90 |
Add the following note: in Amber 6, igb=3 was used to specify a distance-dependent dielectric; in Amber 7, this can be accomplished by setting eedmeth=5 in the &ewald namelist (see p. 100). |
|
p. 104 |
The next-to-last weight for n=9 should be 0.15617 (not 0.15167). |
|
p. 111 |
The first paragraph states, incorrectly: "By default...all input is taken from the standard input file". Actually, we recommend that none of the input for sections 4-7 be on the standard input file; you should should place all restraint information in separate files, and give the filenames here (in section THREE). |
|
p. 135 |
In the example, the eedmeth variable is shown in the &cntrl namelist; actually it must be inside the &ewald namelist. |
|
p. 200 |
The tstep argument to the correlation command is not working properly, and for now should always be set to 1. |
|
p. 252 |
The basic literature reference for the quasih procedure is: R.M. Levy, M. Karplus, J. Kushick, and D. Perahia, Macromolecules 17, 1370-1374 (1984). The -novec flag inhibits outputting of eigenvectors; otherwise, they are written to the file given by the -v flag. The name of the file with the masses is given by the -m flag. |
|
p. 298 |
After CN2, the third sentence should read: "In the case of the dihedrals, if the fourth atom is negative, this implies an improper torsion, and if the third atom is negative, this implies that end group interactions are to be ignored. |