Dear Steve,
> Because of the atom order in Gaussian '03, I'm forced to use ANTE
> RED in non-automatic mode and select OFF for atom order. The p2n
> file I get back looks fine.
this is correct...
to be able to use the Ante_R.E.D. 2.0 automatic mode (the fastest
approach) this means:
- you use Ante_R.E.D. 2.0 to generate the .p2n + .com files; you use
the .com file generated by Ante_R.E.D. on your machine to generate the
corresponding QM .log file in agreement with the atom order defined in
the P2N file; you can then upload the .log + p2n files as inputs for
R.E.D. IV (In "Mode 2" (the default), R.E.D. is executed using the
$OPT_Calc = "Off" & $MEP_Calc = "On" variables)
- or you use only the P2N file as input and let R.E.D. IV performs the
geometry optimization step (In "Mode 1", R.E.D. is executed using the
$OPT_Calc = "On" & $MEP_Calc = "On").
See
http://q4md-forcefieldtools.org/REDS/popup/popredmodes.php
then if you use R.E.D. IV to perform the geometry optimization you are
sure you use Opt=Tight and use the ModRedundant mode only if
required... for DNA nucleosides, I do not think geometrical
constraints are required during geometry optimization. I will forward
to your personal email information about geometrical constraint
implementation in the P2N file for Gaussian.
> (If I use automatic mode, where H's must follow their attached heavy atoms,
Yes
> the p2n visualization shows a C=O double bond between C5' and O5' in 2'-dA)
visualization? do you mean in the java applet?
double bonds do not exist in MM: this is just an interpretation of the
program used to visualize a structure; see the P2N and mol2 files
generated by Ante_R.E.D. and R.E.D.: only simple bonds are generated...
then a simple, double or triple bonds is only represented in the force
field by a different force field constant.
> When I submit that p2n file and the G'03 output to RED, the p2n
> visualization again comes back with C=O bond, but I do get charges
> and the .pdb and .mol2 visualizations show the (expected) H-O and
> C-O single bonds.
only matters the atom order and the presence or absence of a given
bond in the P2N file: the bonds in a mol2 file are these defined in
the corresponding P2N file.
> My question is: does it matter that the p2n file I get back from RED
> shows a C=O? Does that affect the charge calculation? Can I trust
> the pdb and mol2 files?
No
see the P2N and mol2 files: only simple bonds are generated
regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 08 2012 - 10:30:03 PST
