From: Yu Fang (y.fang2_at_ugrad.unimelb.edu.au)
Date: Thu Dec 10 2009 - 05:08:03 CST

Hi Sarah

Your suggestion got me a bit further, but I have a few more questions:

> after the QM calculations, save them into a pdb file. i call it qm.pdb.
-which QM calc are you referring to, the geometry optimization, the single
point calc or the co-ordinate transformation (is this calc required for
parameter generation)?

I have just tried what you suggested with the geometry optimization qm log
file as qm.pdb (saved the last frame of the log file as a pdb using vmd).

In the internal co-ordinates screen, I reassigned atom types and vdw
parameters, then autogenerated internal co-ordinates, all ok so far. Most
of the parameters are loaded from par_all27_prot_lipid.inp.

But there are a few dihedrals for which there are no parameters inside
par_all27_prot_lipid.inp. I would like to use the QM calculation for these
so I assume I highlight this dihedral and click "Use QM Parameters". When
I do this though, the parameters do not appear (force coeff, multiplicity
and phase constant are blank) so when I exit this screen and try to write
the parameter file I get an "expected floating-point number but got """
error.

If instead I make up the missing dihedral parameters and append it to
par_all27_prot_lipid.inp manually, so that clicking on autogenerate
internal co-ordinates (and use force field parameters) button will fill up
all the parameters automatically, then I can successfully write out the
parameter file. But of course this is useless since it is just copying out
what I have done manually. What I need is to get these from the QM calcs.

I suspect that what I need to do is to complete the co-ordinate transform
gaussian job and load the log file, after which in the internal
co-ordinate screen, clicking on "Use QM parameters" or "Recalculate
parameters from Hessian" would compute the parameter from the log file.

Please let me know what you think, and whether I am doing the correct steps.
Thanks,
Yu

> Dear Yu Fung,
> i had the same problem. it's what i did and worked after guidance of
John
> and Peter:
> after the QM calculations, save them into a pdb file. i call it qm.pdb.
then
> choose paratool. choose your initial pdb file as parent molecule and
your
> qm.pdb file as your base molecule. then do whatever you had done before:
1- choose atom types and assign VDW parameters (if some vdw parameters
for
> the atom types are missed, you'll face errors in the next steps) 2-
autogenerate all the coordinations.
> now you can save your topology and parameter files in the file menu. if
you've transferred your ligands from another software, get sure that all
> your atoms have a unique name. otherwise change the name of each atom
and
> start the process (qm and paratool) by your new pdb file.so your ligands
won't break down in the next step like my molecules!
> if you have just one single ligand you can go to autopsf generator and
remove the top_all27_prot_lipid.inp file and replace it with the
topology
> file you have written in the previous step. now you shouldn't have any
problem.
> if you're using the ligand beside any other biological molecule, you should
> (if you're using windows as i guess) add the components of your topology
file to top_all27_prot_lipid.inp in wordpad and save the file in *.txt
format. so this file should be used instead of the original *.inp file.
hope it works correct and accurate for you. i tried to mention all the
details to prevent errors.
> Enjoy life,
> Sarah
>