Dear Holger and others,
What happened was that if running GB and NM together, mm_pbsa.pl assigns
igb=4 to NM minimization also, which slows the process way down!
Now I run MM/PB/GB/SA and NM separately. And it works fine.
Thanks,
Chenglong Li
attached mail follows:
Chenglong Li wrote:
>
> Dear AMBER users,
>
> In MM_PB/GB/SA analysis, I have problems with both un-reduced and
> reduced complex approaches:
>
> 1) With the whole system, the igb=4 minimizations are too slow because
> of
> the large size of the protein.
I don't quite see where the minimizations come into play? The basic idea
of MM_PB/GBSA is to generate an ensemble of structures and then to use
these structures (w/o minimization) for calculation of the MM, PB/GB,
and SA part. In the case of estimating vibrational entropies by normal
mode analysis (the only place where I would use a minimization), a
reduced number of snapshots can be used without too much affecting the
precision of the results.
Best regards
Holger
> If I change to cut=16 from without
> cutoff,
> it converges much faster so I can use much fewer maxcyc.
> But I don't know if it's a good choice because a) the energies are
> different;
> b) MM may be off quite a bit. Even if deltaG may not got affacted much
> due to
> cancelling-out, still std.dev. may be larger than it should be.
>
> 2) If I cut a reduced system out centered around the ligand, the new
> complex
> contains a lot of short pieces of individual peptide strands. The
> re-creaction of
> the receptor and complex prmtops introduces some new H in the
> N-terminals and
> new OXT in the C-terminals. These new atoms of course are not in the
> original
> snapshots taken out of MD trajectories. It looks that you cannot get a
> prmtop
> with leap without these new atoms because leap would complain the large
> distances between the terminal atoms. I wonder how to get a workable
> reduced
> complex model.
>
> Ant solution is welcome.
>
> Thanks,
> Chenglong Li
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
--
++++++++++++++++++++++++++++++++++++++++++++++++++
Dr. Holger Gohlke
J.W. Goethe-Universität
Fachbereich Biologie und Informatik
Institut für Mikrobiologie
Marie-Curie-Str. 9
60439 Frankfurt/Main
Germany
Tel.: (+49) 69-798-29503; Fax: (+49) 69-798-29826
Email: gohlke.bioinformatik.uni-frankfurt.de
URL: http://www.rz.uni-frankfurt.de/~hgohlke
++++++++++++++++++++++++++++++++++++++++++++++++++
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Thu Nov 20 2003 - 18:53:00 PST
