On Tue, May 13, 2014 at 4:47 PM, MURAT OZTURK <murozturk.ku.edu.tr> wrote:
> Thanks for the response.
>
> Yes, I am sure that I am using ntb = 0. I also know that it works prior to
> adding ions, so adding ions is what breaks it.
>
Yikes... Don't do this ;). If you plan on running with explicit solvent,
do this off the bat. (pmemd.cuda doesn't even support vacuum dynamics).
The best way to add ions to a system via tleap is actually to add the
solvent first and then replace random solvent molecules with ions (at least
IMO -- YMMV). So something like this:
system = loadPDB you_system.pdb
ions = loadAmberParams frcmod.ionsjc_tip3p
solvateOct system TIP3PBOX 15.0
addIonsRand system Na+ 0
saveAmberParm system your_system.parm7 your_system.rst7
The first line loads your system. The second loads the ion parameters (as
Ilyas mentioned, the "ions=" part is optional). The third solvates the
system with TIP3P waters in a truncated octahedron with at least a 15
Angstrom buffer around the solute. The fourth replaces enough waters with
sodium ions to neutralize the system. The last saves your topology file.
(Of course this assumes a default leaprc file, which you mentioned you had.)
> Interestingly, the last two lines of the inpcrd of the protein only system
> reads
>
> 82.4450000 14.3940000 -13.6870000 81.7180000 15.3700000 -13.4430000
> 83.6048830 14.5521392 -13.3126966
>
> whereas the inpcrd with ions reads
>
> 86.3470000 12.2630000 -15.2900000 87.2459461 11.6555756 -15.3950273
> 86.5023142 13.2275008 -15.7734415 86.1831237 12.5406017 -13.8148107
> 87.2007
> This does not look correct to me. But it is purely generated by leap.
>
Looks like you got clipped with a bug we recently found (LEaP occasionally
truncates files). An update should be posted tomorrow morning (EST in the
USA) -- here is the relevant thread on the mailing list describing the bug
http://archive.ambermd.org/201404/0532.html (there is a proposed patch
attached there that you can download and apply if you don't want to wait).
Why are there 6 columns in the first place? I'd expect 3 columns *
> numberOfAtoms rows. Forgive my gromacs inclination, is this the amber way?
>
Yes -- this is the amber way. It is the x, y, z coordinates of two atoms.
> (CRD spec states FORMAT(6F12.7), which is again cryptic to me.)
>
This is a Fortran format statement. It indicates that there are at most 6
Floating point numbers per line, each with 7 decimal points that take up
exactly 12 characters. Amber is traditionally a Fortran code-base
(exclusively back in the early days of Amber 3 to 4, although it is FAR
more diverse now). You can still see strong remnants of this Fortran
heritage everywhere.
And in the last one, there is a row with a single column. Is this supposed
> to happen?
>
No. It's a bug that should hopefully have a fix posted soon (see above).
>
> Regards,
>
> Murat
>
>
>
> Here is the full(ish) output including the error. :
>
> [snip]
>
> --------------------------------------------------------------------------------
> 1. RESOURCE USE:
>
> --------------------------------------------------------------------------------
>
> | Flags:
>
> | NONPERIODIC ntb=0 and igb=0: Setting up nonperiodic simulation
> At line 842 of file ew_box.F90 (unit = 9, file = 'inpcrd')
> Fortran runtime error: End of file
>
You got hit by the leap bug -- the file was truncated.
Hope this helps,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 13 2014 - 15:30:02 PDT
