AMBER Archive (2008)

Subject: Re: AMBER: Nmode segmentation fault

From: James Thomas (james.protein_at_googlemail.com)
Date: Tue Mar 25 2008 - 12:22:43 CDT


Hi Amber community

Test protein: 1382 atoms (successful)
Interest protein: 5780 atoms (segmentation fault error)

I did not do any CG minimisation before starting NR.

I look forward to your advice on how to solve the segmentation fault problem

Many thanks!

On Tue, Mar 25, 2008 at 4:34 PM, David A. Case <case_at_scripps.edu> wrote:

> On Tue, Mar 25, 2008, James Thomas wrote:
> >
> > I am running a nmode program using amber9 package. I tried using a
> smaller
> > file as a test (the prmtop size: 566999 & the inpcrd: 50451). My
> interest
> > protein has this (the prmtop size: 2307687 & the inpcrd size:210978).
> > However, I got the error message of segmentation fault few minutes after
> I
> > started the run.
>
> It is much more helpful to know how many atoms you have in each case; the
> size of the prmtop file itself is not very informative.
> >
> > My test run using test protein is successful to complete 100 steps of
> Newton
> > minimization.
>
> This sounds suspicious: Newton Raphson should generally converge in much
> less
> than 100 steps. (You should have done a lot of CG minimization before
> starting NR.)
>
> >
> > I also refer this : http://amber.scripps.edu/bugfixes/6.0/bugfix.28, but
> not
> > sure on how to modify the alloc.f file.
>
> This is about amber 6(!). Should be completely irrelevant for Amber 9.
> >
> > Is that because my protein of interest is too big?
>
> See above...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu