# Re: [AMBER] force field parameter and resp charge calculation

From: FyD <fyd.q4md-forcefieldtools.org>
Date: Wed, 30 Jun 2010 09:15:50 +0200

Dear Colvin,

Charge fitting is different from charge averaging!
See
http://lists.q4md-forcefieldtools.org/wws/arc/q4md-fft/2010-05/msg00024.html

Shortly:
- When you do charge averaging for X orientations (and/or X
conformations), you derive X charge sets based on X single orientation
charge fits and then you average the X charge sets into a single one
- When you do charge fitting based on X orientations, you derive a
single set of charges based on X MEPs/X orientations

Usually, in RESP charge derivation charge averaging is not used.

You can use whatever number of conformations/orientations you want in
RESP charge derivation. However, I wonder if 24! is really useful...

regards, Francois

> Dear Francois,
>
> by the way, can i just average the charges over the 24 conformations? Pls
>
> Thank you very much.
>
> On Wed, Jun 30, 2010 at 2:21 PM, FyD <fyd.q4md-forcefieldtools.org> wrote:
>
>> Dear Colvin,
>>
>> I forgot a point:
>> If you are interested in molecular orientation, in R.E.D.-III.4 the
>> rigid-body reorientation algorithm has been rewritten allowing to
>> study the impact of translation versus rotation (besides the
>> re-orientation procedure originally defined in R.E.D. I).
>> See http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#Translate
>> http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#Rotate
>> versus
>> http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#QMRAALGO
>> http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#RBRAALGO
>>
>> Thus, any user can demonstrate by herself/himself the origin of the
>> problem of charge reproducibility.
>>
>> regards, Francois
>>
>>
>> > Dear all.
>> >
>> > I was trying generate force field parameter for a molecule with 134 atoms
>> > and 24 different orientations/conformations using GAFF, following the
>> > tutorial provided from http://ambermd.org/antechamber/example.html and
>> > ambertools user manual.
>> >
>> > These are the steps used:
>> >
>> > 1. antechamber -i *.mol2 -fi mol2 -o *.com -fo gcrt (* = number
>> corresponds
>> > to 24 different orientations)
>> > 2. run gaussian with the default input file generated from antechamber
>> > 3. espgen -i *.log -o *.esp (extract out the esp)
>> > 4. cat *.esp > all.esp (merge all the esp files into one)
>> > 5. respgen -i *.ac -o step1.respin1 -f resp1 -n 24
>> > 6. respgen -i *.ac -o step2.respin2 -f resp2 -n 24
>> > 7. resp -O -i step1.respin1 -o step1.respout1 -e all.esp -t qout_stage1
>> > (default resp input file from respgen)
>> > 8. resp -O -i step2.respin2 -o step2.respout2 -e all.esp -q qout_stage1
>> -t
>> > qout_stage2 (default resp input file from respgen)
>> > 9. antechamber -i 1.ac -fi ac -o resp.ac -fo ac -c rc -cf qout_stage2
>> >
>> >
>> > My question is:
>> >
>> > 1. Are the procedure above correct?
>> >
>> > 2. if i use the first 3 molecules and above and concatenate the
>> correspond
>> > esp files for resp calculations, the charges are all the same.
>> > for example, cat 1.esp 2.esp 3.esp > first3.esp --> respgen with -n 3
>> --->
>> > do resp calculation using first3.esp and read qout_stage2 to 1.ac (same
>> > apply to 1-4molecules, 1-10molecules and 1-24 molecules, the charges are
>> the
>> > same --> charge set A)
>> >
>> > if i use only 2 molecules (molecule 1-2), the charges are different
>> compared
>> > to the above case -->charge set B.
>> >
>> > if i use the last 3 molecules and below, the charges are the same as well
>> > but are different with case 2.
>> > for example, cat 24.esp 23.esp 22.esp > last3.esp --> respgen with -n 3
>> --->
>> > do resp calculation using last3.esp and read qout_stage2 to 1.ac (same
>> apply
>> > to 24-21molecules, 24-15molecules and 24-1 molecules, the charges are the
>> > same --> charge set C)
>> >
>> > I am confuse with this...is there anything wrong in the procedure? it
>> seems
>> > that the charges generated somehow related to the sequence of the
>> concatenate
>> > esp files.....is it?
>> >
>> > or shall i just do resp calculation for separate 24 orientations and
>> average
>> > the charges generated?
>> >
>> >
>> > Thank you very much.
>> >
>> > Regards,
>> > colvin
>>
>> http://q4md-forcefieldtools.org/FyD/
>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>