AMBER Archive (2005)

Subject: Re: AMBER: nmode with ibelly

From: Martin Sippel (masi_at_nmr.mpibpc.mpg.de)
Date: Tue Jun 07 2005 - 05:11:44 CDT


('binary' encoding is not supported, stored as-is) Dear dac,

Thanks for your answer. After modifying sizes2.h and recompiling nmode, employing the ismem = 1 option, there are no
error messages anymore. But am I right that with ibelly = 1 option the entropy calculation doesn't work (there ist this
message: "Thermo analysis not supported for belly calc.") ?! How can I define that only certrain residues should be considered
for the entropy calculation to speed it up?

Regards,

Martin

>>
>> using nmode in amber8, I want to do some entropy calculations on a
>> protein-ligand-complex. The system is pretty large, hence I employ the
>> ibelly-option to fix the major part of my protein, but even with only two
>> residues allowed to move, nmode still gives the "segmentation fault" error
>> message, nothing else. I have 2 gigs of memory, so that should be no
>> problem.
>
>Amber's nmode does not currently reduce the space requirements based on the
>belly specification; see lines 90-100 of $AMBERHOME/src/nmode/nmode.f.
>
>I am not quite sure why you are getting segmentation faults. The allocation
>statements (lines 104 and 107 of the above file) should exit cleanly with
>a useful error message if enough space is not available. We need to figure
>out why this doesn't happen.
>
>...regards...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