Dear Tao,
the generic X-CM=CM-X dihedral FF parameter:
X -CM-CM-X 4 2.66000000e+01 180.0 2.
-> there are four paths
The specific CA-CM=CM-CA dihedral FF parameter:
CA-CM-CM-CA 1 6.65000000e+00 180.0 2.
-> there is a single path
you should also find:
HA-CM-CM-HA 1 6.65000000e+00 180.0 2.
CA-CM-CM-HA 1 6.65000000e+00 180.0 2.
2 CA-CM-CM-HA + 1 CA-CM-CM-CA + HA-CM-CM-HA = X -CM-CM-X
both FF representations (specific vs generic) fortunately lead to the
same energy value; however PyRED generates only specific FF parameters
and transforms the Amber generic FF parameters into specific ones
before to print them in the frcmod.known/correspondence files. I think
you misunderstand the dihedral profile or I misunderstood your
question ;-)
See
http://ambermd.org/doc6/html/AMBER-sh-19.3.html#sh-19.3
(the problem is that in this documentation the figures are missing;
try to find a paper version)
my feeling is that if the double bond is 'trans' in your input
structure (if I use your terminology; may-be 'E' vs 'Z' would be more
correct) it will remains 'trans' during MD at 300K (or after geometry
optimization). similarly if the double bond is 'cis' in your input
structure it will remains 'cis' at 300 K.
regards, Francois
> Thanks a lot for the detailed instructions with the PyRED usage and atom
> types. I've submitted a new job with pre-defined atom types, and everything
> looks ok. Now I am facing a new problem with the dihedral parameter of the
> CA-CM=CM-CA. Theoretically, it should has two configurations, one is
> "cis", and another is "trans". However, when I looked into the existed
> dihedral parameter for CA-CM-CM-CA in frcmod.known, the only one could be
> found is
>
> CA-CM-CM-CA 1 6.65000000e+00 180.0 2. adapted from
> parm99.dat i.e X-CM-CM-X 26.6/4
>
> Maybe it is a dumb question, does that mean it only defined the "trans"
> configuration? Should I worry about it? I've run two tests with two
> different start configurations (cis and trans), but neither of them
> transitioned to another.
> 2015-03-07 0:20 GMT+08:00 FyD <fyd.q4md-forcefieldtools.org>:
>
>> Dear Tao,
>>
>> Yes I know we have to update our dictionary of atom types related to
>> the CM vs CD atom types. We are correcting a series a limitation; On
>> our todo list: we modify the web server, the PBS script and implement
>> new features in PyRED almost every day...
>>
>> R.E.D. Server Dev. is saturated these days: we have to open more nodes
>> also - uff...
>>
>> In your case:
>> - Run a 1st PyRED job
>> - Check the atom types as you did
>> 1) if you like the generated atom types look for missing FF
>> parameters (frcmod.unknown file)
>> and use the 'Re_Fit mode to provide these missing FF parameters
>> (frcmod.user file)
>> i.e. http://q4md-forcefieldtools.org/Tutorial/Tutorial-4-demo1.pdf
>> -> END: ready to validate the FF generated by PyRED.
>> 2) if you do not like the generated atom types
>>
>> grep MOLECULE1-ATMTYPE
>> Data-R.E.D.Server/Data-Default-Proj/Project-out.config >>
>> ./Project.config
>> -> and modify the listed atom types according to what you think
>>
>> mv Mol_red1.pdb Mol_red1.pdb.origin (old atom order)
>> cp Data-R.E.D.Server/Data-Default-Proj/Mol_red1-out.ent ./Mol_red1.pdb
>> (to be sure to have a correct atom order; optimized geometry)
>>
>> rerun the job as you did the 1st time, check the atom types as you did
>> again (should be ok this time) and look for missing FF parameters
>>
>> --
>>
>> in your case I would start by defining the CA atom types for all the
>> sp2 carbon atoms in the two aromatic rings; and may be replace CD by
>> CM for that double bond between these 2 rings...
>>
>> regards, Francois
>>
>>
>> > Thanks a lot for providing such invaluable information. I've read the
>> > tutorial and run two jobs on the R.E.D. Develoment server. However,
>> > the atom typing of one looks a little strange, so I am looking for help
>> for
>> > this molecule (see attached picture).
>> >
>> > Generally, this molecule was two aromatic ring linked by ethylene. The
>> > R.E.D server could correctly calculate the partial charge for it, but the
>> > assigned atom types surprised me. The two carbon atom from ethylele was
>> > assigned atom type "CD", which was mentioned in parm99.dat as "sp2 C atom
>> > in the middle of: C=CD-CD=C". But, in our case, the two carbon atoms was
>> > connected by double bond, not single bond. I also found the description
>> of
>> > "CD" from Wang et al. 2000, which said that,
>> > "Considering that the inner two carbons of butadiene are not pure sp2
>> > carbons, and the bond lengths between them are slightly longer than a
>> pure
>> > single Csp2?Csp2, a new atomic type (CD) can be introduced. "
>> >
>> > So, I am a little confused. Is CD the proper atom type for those two
>> > carbons? Or do you have any experience/examples on such conjugated
>> systems?
>>
>>
>> > 2015-03-03 17:31 GMT+08:00 FyD <fyd.q4md-forcefieldtools.org>:
>> >
>> >> Dear Tao,
>> >>
>> >> See http://q4md-forcefieldtools.org/REDServer-Development/
>> >> and its tutorial:
>> >> http://q4md-forcefieldtools.org/Tutorial/Tutorial-4.php
>> >>
>> >> and particularly:
>> >> http://q4md-forcefieldtools.org/Tutorial/Tutorial-4.php#17
>> >> http://q4md-forcefieldtools.org/Tutorial/Tutorial-4.php#18
>> >>
>> >> in agreement with:
>> >> http://onlinelibrary.wiley.com/doi/10.1002/jcc.540161106/abstract
>> >>
>> >> regards, Francois
>> >>
>> >> > I am doing simulation on system that with various compounds bonded to
>> the
>> >> > N-terminal of certain peptide, like shown in the following picture.
>> Now I
>> >> > am facing the problems with creating unit for such N-Cap residue.
>> >> >
>> >> > ?
>> >> > 1. Should I treat R1 combined R' as a N-Cap residue or just the R'
>> only?
>> >> > 2. In any case above, is there any method available to assign correct
>> >> > partial charge and atom type for such non-standard residue?
>>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Mar 08 2015 - 10:00:03 PDT
