[gmx-users] RE: Define proper set of topology parameters for new molecule in different FF.

[Justin A. Lemkul]

Andrey Frolov wrote:
>>> I want to construct TOPOLOGY of many different molecules using different
>>> Force Fields (FF).
>>>
>>> If i want to describe , for example, METHYL ACETATE in OPLS-AA force 
>>> field
>>>
>>>          O             Hc5
>>>         ||             |
>>>          C-------O1----C2----Hc4
>>>          |             |
>>>    Hc3-- C1--Hc2       Hc6
>>>          |
>>>          Hc1
>>> 1) I'll find all ATOM TYPES in oplsaa, let`s imagine that they are like
>>> in the image above.
>>> 2) then i describe all BONDS and PAIRS   3) then i should define BOND 
>>> ANGLES
>>>     but how much angles should i describe?
>>>     PRODRG and topologies in share/top give all possible angles,
>>>     but is there overestimation of FF energy in such cases, as:
>>>           angle O-C-C1
>>>           angle O-C-O1
>>>           angle O1-C-C1
>>>       3 adjoining angles in planar group?
>>> 4) then i should define PROPER and IMPROPER DIHEDRALS.
>>>     Should i define all possible DIHEDRALS as it is made in GMX
>>> share/top/urea.itp?
>>>     Or should i define only those DIHEDRALS that are exist in chosen FF?
>>>         Are there any rules how to chose these parameters?
>>>
>>> For the first view it seems that you should chose these parameters like
>>> they were parametrized in chosen FF.
> 
>>> => the "topology graph" (gmx topology  without  all FF values) is
> 
>> >>different for different FF?
>> >>     Am i right?
>> >>     And in order to describe the molecular TOPOLOGY in CERTAIN FF we
>> >>should know the special "rules" for THIS FF,
>> >>=> mail to author of FF or read initial papers.
>> >>     Am i right?
>> >>
>> >>The initial papers do not contain CLEAR description of how to chose
>> >>proper set of parameters.
>>   
>  >> I`m interested in GROMOS, OPLS-UA, OPLS-AA, AMBER-UA, AMBER-AA FF,
> 
>> >>MMFF94.
>>   
>  >> So, if you know any rules or useful links i'll be very thankful.
> 
>> >>I`m very sorry if my questions are mentioned already, but i did not 
>> find
>> >>the answer in mailing-list.
>> >>Kindest regards,
>>   
> 
>> You should describe the values of those parameters you want either
>> constrain or simulate with intramolecular parameters. Evidently, you
>> need a set of values completely determining the geometry of your
>> molecule. However, you make want some atom groups to rotate during the
>> simulation, then you need not to give the respective numbers.
> 
>> For an example just see the existing FF provided with GROMACS..
> 
> Dear Dr. Chaban,
> 
> Thanks a lot for your reply.
> I`m very sorry but it seems that you did not properly understand the 
> goal of the letter.
> 
> 1) The aim is to create a topology in certain FF
> that`s why we should follow the methodology of parametrization of this 
> FF (to my mind).
> |=> the "topology graph" (gmx topology  without  all FF values) is not 
> FF independent.

I don't know what use a topology with no parameters would be...a list of the 
atoms?  That would be a structure file, and so, yes, it would be independent of 
the force field.  You could, for example, make a skeleton topology that has all 
the sections (atoms, bonds, etc.) and use it as a template for all the force 
fields, because the .top structure is fairly generic.  See Chapter 5 of the 
manual for more details.


>     Am i right?
> 2) To describe exact 3D structure of molecule we should    define 2*N-5 
> angles (bond angles or/and dihedrals) (N-number of atoms),
>    besides of all bonds.
>     (that`s my empirical formula).
>    In gmx provided topologies there is no possibility to understand
>    what is the final topology of residue (RTP file: there are only bonds
>        and improper dihedrals, what about others: bond angles, and 
> proper dihedral?)

That depends on the force field.  In the .rtp entries for, say, OPLSAA, no 
angles are specified, and so you are correct.  However, in the generation of the 
topology, you will find that angles are present!  This is because angles are 
defined within ffoplsaabon.itp based on atom naming.  If you are manually 
generating a topology, you will have to include these angles.

If you look, for example, at ffG53a6.rtp, you will find explicit inclusion of 
all the bonded terms (bonds, angles, impropers, dihedrals); such is the format 
of the Gromos force field files.

It might be helpful to run through some tutorial material using a simple system 
(like a protein), using each of the force fields you are interested in, to see 
how parameters are generated in the topology by pdb2gmx.

For more information about pairwise interactions, etc., there is no substitute 
for reading the original literature (and subsequent citations and updates to) 
for each force field.

-Justin

>       share/top/urea.itp: there are all possible bond angles
>  H            8 proper dihedrals (max value)
>  |            3 improper dihedral (max value)  N--H
>  |
> O=C--N--H
>     |               H
> The set of parameters is more then enough.
> 
>    share/top/benzamide.itp: (that is not realy benzamide if we`ll see 
> WIKIpedia)
>     H            there are set of parameters that is greater that it
>     |             should be to define 3D structure, but less then the 
> max value.
>     N--H          __  |
> /O \_C--N--H
> \__/    |
>        H
> 
> Again the questions are the same as in the first letter.
> 
> Thanks a lot.
> I`m looking forward to your reply.
> Andrey Frolov.
> 
> Institute of Solution Chemistry, RAS.
> Ivanovo, Russia.
> 
> 
>     
> 
> 
> 
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the www 
> interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 

-- 
========================================

Justin A. Lemkul
Graduate Research Assistant
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================