Dear Olga,
> I am trying to use RED III software to get resp charges for anionic
> fragments to use them with GLYCAM06.
> It seems that there is no such an option (model) with the HF/6-31++g(2d,2p)
> in the RED-vIII.5.pl script.
> So it would be the same option as option 4 but with the different basis
> set:
> HF/6-31++g(2d,2p) CHELPG algo., 1 stage RESP fit qwt=.01.
>
> Is there any chance that you would add additional option for GLYCAM06 to
> the RED III package?
This has been discussed in the Amber mailing list some times ago; I
think this is better writing this type of question in the q4md-fft
Mailing list .
http://lists.q4md-forcefieldtools.org/wws/info/q4md-fft
or in the Amber mailing list and not .
contact_at_q4md-forcefieldtools.org.
Thus, you will get the opinions of the GLYCAM, AMBER and q4md-fft developers.
In R.E.D. III.x, the charge model used to match the GLYCAM approach is
RESP-C2.
See
http://q4md-forcefieldtools.org/REDS/popup/popkeyword.php
i.e. -5- RESP-C2
HF/6-31G*//HF/6-31G* - CHELPG algo. used in MEP computation - 1 stage
RESP fit qwt=0.01
We are obviously aware that double zeta basis set with diffusion
functions should be better to account for molecules with hydrogen
bonds (I mean only in the geometry optimization step). However, in
R.E.D. we only propose original approaches used in Amber, glycam,
(hope soon other) force fields. Thus, there is no mistake and this is
normal that HF/6-31G* is used in geometry optimization even for
monosaccharide units. My understanding of using HF/6-31G* in gas phase
in geometry optimization for monosaccharide, amino-acid and nucleoside
residues is only to generate _canonical_ structures to be involved in
_empirical_ force field development [while structures should be
obviously different in condensed phase even when using
HF/6-31++g(2d,2p)]. Structures with canonical intra-molecular hydrogen
bonds are also discarded from MEP computation when using HF/6-31G*
(see examples in R.E.DD.B.).
That being said you can obviously adopt your own approach and modify
the source code of R.E.D. so that geometry optimization is used with
HF/6-31++g(2d,2p) instead of HF/6-31G*.
In R.E.D. III.5, see the file handle JOB1_FILE lines 1656-1667 for
Gaussian; simply replace HF/6-31G* by HF/6-31++g(2d,2p) where this is
needed.
For GAMESS and Firefly, the modifications to be performed are a little
before in the source code; still in the JOB1_FILE file handle.
Obviously, HF/6-31G* MUST be used in MEP computation.
I hope this helps...
regards, Francois
PS
See the bug fixes
http://q4md-forcefieldtools.org/RED/popup/bugs-III.5.txt...
We released the III.51 tools to help users yesterday...
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Feb 25 2012 - 00:30:02 PST
