AMBER Archive (2004)

Subject: Re: AMBER: Using pbsa

From: Chenglong Li (chenlong_at_scripps.edu)
Date: Tue Apr 20 2004 - 18:47:34 CDT

Hi Thomas,

It is a bug as Ray tested out. Just downgrade the bomb to a warning in
pb_force.f:

<<<<<<
write(6, '(a,3i4)') 'PB Bomb in epsbnd(): No neighbor found .... '
call mexit(6,1)
>>>>>>
nwarn = nwarn + 1

and re-compile.

Good luck,
Chenlong Li

-------------------------------

Thomas Steinbrecher wrote:

> Dear Amber users,
>
> I am trying to do MM-PBSA calculations on a protein (~25kDa) ligand
> complex, using amber8 on Redhat Linux. I use mm_pbsa.in as provided in
> the examples, except that I had to set radioopt to 0 because my ligand
> contains gaff-atomtypes.
>
> When I let mm_pbsa.pl run over my snapshots, the pbsa calculation fails
> always at the same complex snapshot.
>
> The last line in the pbsa*.out contains this error message:
>
> PB Bomb in epsbnd(): No neighbor found for exposed boundary grid 22
> 60 16
>
> The snapshot where the calculations fails is not visibly different from
> the one before (rms <1.5A)
>
> Can anyone tell me what this error means or how to fix it?
>
> Kind regards,
>
> Thomas
>
> -----------------------------------------------------------------------
> 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