AMBER Archive (2003)

Subject: Re: saveAmberParm problems in xleap

From: Ioana Cozmuta (ioana_at_nas.nasa.gov)
Date: Tue May 06 2003 - 15:54:49 CDT

Hi,

I did use tleap (text version for large systems is a bit better from the
point of view of speed) to solvate a protein (2056 residues,about 35000
atoms, solvated the system has 200,000 atoms) and I had some difficulties
in the beginning but they were related to the way I defined the sequences
or some other small details that sometimes you do not think of. I found
out that actually if you use your own sequence and then use
loadPsbUsingSeq you reduce the amount of errors that tleap may generate.
About InsightII, I did load the same structure file in InsightII and it's
tricky but in the end you can find a way around to save the pdb structure
so that it is compatible with leap. Sometimes I found that it is a good
idea to remove the hydrogens and let tleap add them back. But then you
have to make sure to check the protonation states of your aminoacids.
In my case I used an SGI station(Octane dual) so maybe your problem is -as
Guanglei suggests- machine specific.

Hope this helps,
IOana

 On Tue, 6
May 2003, Guanglei Cui wrote:

> Hi,
> I've used xleap to create solvated system that is as large as 70,000 atoms
> and didn't have any crashes with either amber6 or amber7. Could it be
> machine-specific? I have a dual PIII 800/Linux.
>
> On Tuesday 06 May 2003 04:14, Wayne Dawson wrote:
> > Dear Amber users,
> >
> > I am trying to run a 20000 atom simulation (protein in
> > explicit water w/ PBC). However, whenever I set up the protein
> > + water complex, xleap dies when it tries to create the
> > *.crd file.
> >
> > > saveAmberParm prtn_wtBox prtn_wtBox.top prtn_wtBox.crd
> >
> > .....*** He's dead Jim ***.....
> >
> > It must be an array dimension problem in the initial
> > configuration of the program, but how can I enlarge the
> > array?
> >
> > I noticed that someone posted a similar problem last year in the
> > amber reflector,
> >
> > http://structbio.vanderbilt.edu/archives/amber-archive/2002/0489.phtml
> >
> > However, there was no answer to this post. That post did provide a
> > "work around" (i.e., use tleap), but has this problem been solved?
> > What configuration files should I fix if so?
> >
> > I gather from
> > http://structbio.vanderbilt.edu/archives/amber-archive/2003/0685.phtml
> > and some other posts that I can (at least in principle) do
> > calculations of this size.
> >
> > Thank you.
> > Wayne
>
> --
> Guanglei Cui
> Dept. of Chemistry
> SUNY at Stony Brook
> Stony Brook, NY 11790
>
>