AMBER Archive (2008)Subject: Re: AMBER: REDvIII_I_lack_of_charges_in_mol2
From: FyD (fyd_at_q4mdforcefieldtools.org)
Date: Wed Oct 22 2008  10:15:15 CDT
Quoting Karol Kaszuba <karolkaszub_at_gmail.com>:
> I am trying to derive RESP A1 charges. I am using Gaussian.
> RED fnishes normally without any errors, (RED.log is absolutely OK) however
> the .mol2 file doesn't have any charges  all columns are 0.000.
R.E.D. should have detected this problem:
See the new bug fix on the R.E.D. web site.
> I started digging and I figure out that:
>
> 1) Instead off 16 files after RED execution I have only 14.
> The missing files are:
> punch2_m1
> qout2_m1
>
> 2) In the file: output1_m1 I could see this warning:
> WARNING: Restraints were requested, but the restraint weights were all
> zero,
> then:
> Point Charges Before & After Optimization
>
> no. At.no. q(init) q(opt) ivary d(rstr)/dq
> 1 7 0.000000 nan 0 nan
> 2 1 0.000000 nan 0 0.000000
> 3 6 0.000000 nan 0 nan
> 4 1 0.000000 nan 0 0.000000
> 5 6 0.000000 nan 0 nan
> 6 1 0.000000 nan 0 0.000000
> 7 1 0.000000 nan 0 0.000000
> 8 6 0.000000 nan 0 nan
> 9 7 0.000000 nan 0 nan
> 10 1 0.000000 nan 0 0.000000
>
> Sum over the calculated charges: nan
>
> Statistics of the fitting:
> The initial sum of squares (ssvpot) 21.206
> The residual sum of squares (chipot) nan
> The std err of estimate (sqrt(chipot/N)) nan
> ESP relative RMS (SQRT(chipot/ssvpot)) nan
This is not a problem in R.E.D. The problem is because RESP cannot fit...
I guess this is because your optimized structure is a test case and is
not optimized. I am quite surprised Gaussian is able to SCF converge
to compute the MEP for your structure...
Do not optimized by QM a molecular fragment, but a whole molecule.
Are you sure the total charge is 1 and the multiplicity 2 ?
It think this is because you have a molecular fragment in your QM job.
regards, Francois

The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
to majordomo_at_scripps.edu
