I still feel that my best chances are through divcon. I gratefully accept your
suggestion to go to divcon directly, chapter 7. Checking against QuantumBio web
pages, it seems that the version of divcon in Amber9 is the latest, 4.5.
Antechamber chapter 4 might have been more explicit, telling immediately that
the job will be "Standard Job", i.e., without linear scaling. I was mislead by
the temperature setting, that Amber9 manual lists under the "divide and
conquer" option. After your suggestion above, I examined divcon.in created by
antechamber (precisely by wmopcrt) for divcon for my 500-atoms molecule. Of the
keywords listed:
CARTESIAN AM1 STANDARD DIRECT OPT=BFGS XTEST=0.0001 CHARGE=-2
STANDARD shows that minimization is going on hopelessly with the N^3 scaling
with the number of electrons.
If I understand, appropriate modification of the keywords on top of divcon.in
should make this native divcon input file (divcon.in) suitable to the task. If
so, Antechamber is useful to provide divcon.in. Then, I have to run
linearly-scaled divcon so that it furnishes files for the other antechamber
components to arrive at prmtop and inprcd.
All that assuming that linearly-scaled divcon is able to treat my 500-atoms
molecule as a single residue (I assume that the examples given in divcon manual
for multi-residue molecules simply reflect the prevalent use).
In summary, the situation is now much clearer. Any suggestion about what I have
grossly delineated above would be welcome.
Thanks
francesco pietra
--- "David A. Case" <case.scripps.edu> wrote:
> On Wed, Aug 15, 2007, Francesco Pietra wrote:
> >
> > In my view, the worst about present implementation of divcon in antechamber
> > (Amber9) is lack of keywords DUMP=int PDUMP=int SNAPGEOM as from
> divcon.out,
> > unless they can be set (how?). This is the main need by users of
> antechamber
> > for heavy work. Launching antechamber as in tutorial B6 is risky: if it
> > crashes, all work is lost.
>
> You are really taking the charge derivation part of antchamber where it was
> not designed to go. Even it things don't crash, you would need to examine
> whether or not the charge model one gets for very large molecules is a good
> one.
>
> The "am1-bcc" charge philosophy is to perform a geometry minimization on the
> input molecule before assessing the charges. This makes sense for small
> molecules, where geometry optimization is fast. For large, floppy molecules,
> it probably makes much less sense. Again, this is a scientific question that
> will require patience and research. If you wish to use divcon for this
> purpose, then you should probably run it by hand, not through the antechamber
> front end; then you can set whatever keywords you want.
>
> As others have pointed out, the "divcon" program can perform either "regular"
> SCF calculations, or divide and conquer ones, depending on its input.
>
> ...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
____________________________________________________________________________________
Pinpoint customers who are looking for what you sell.
http://searchmarketing.yahoo.com/
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Aug 19 2007 - 06:07:15 PDT
