[gmx-users] multiple forcefields

[Yiannis]

Hello,
maybe I was not very clear:
I wanted to simulate a fatty acid, so I thought using the lipid.itp of 
Dr. Tielemann and gmx forcefield.
Before my simulation, I tried as a test his DPPC exemple from the web 
page using his topology, gro and lipid.itp and it worked as expected.
But when trying my system I had problems with my fatty acid trying to 
use a similar configuration with gmx and lipid.itp.
So I switched to opls force field, wrote a residue topology in 
ffoplsaa.rtp for my fatty acid choosing manually atom types and using 
pdb2gmx to generate my top file from this topology. So I finally don't 
use the lipid.itp file of Dr. Tielemann and I don't think it can be 
merged with the opls force field as opls is an all-atom and lipid.itp 
uses united atom types (if somebody can confirm this...).
The only think I had to modify is that for esters the atom type OS1 is 
not parametrized in the ffoplsaabon.itp file, so I used some of the OS 
parameters and the parameters from a paper by Jorgensen (J. Comp. 
Chem., vol. 22 (2001) p. 1340). I think the parameters for the OS 
dihedrals in the ffoplsaabon.itp  are the same as those in this paper.
Hope it's clear now.
Ioannis


Le mercredi, 23 jun 2004, à 14:43 Europe/Paris, Derrick Guang Yuh Lee a 
écrit :

> dear yiannis
>
> sorry about my misintereptation of your previous message about the 
> system
> that you are working with, but my question still stands. i am working 
> w/ a
> DPPC system and that is not recognized by the OPLS rtp file, can i 
> simply
> write into the rtp the appropriate data?
>
>
>  - derrick
>
>  derrick Lee
>  faculty of Science
>  Memorial University of Newfoundland
>  k24dgyl at mun.ca / derrickglee at hotmail.com
>
> "a teacher is never a giver of truth - he is a guide, a pointer to the
> truth that each student must find for himself. a good teacher is 
> merely a
> catalyst."
>                                                               - bruce 
> lee
>
> On Tue, 22 Jun 2004, Yiannis wrote:
>
>> Hello,
>> just to add my experience, I tried recently to use lipid.itp from P.
>> Tieleman's site to model a box of polyunsaturated fatty acids using
>> ffgmx force field. While I could use his dppc bilayer topology and run
>> the simulation I had stability problems with my system. Most probably
>> these instabilities (leading to crash after 200 to 400 ps) were due to
>> my topology around the unsaturated bonds. I finally decided to give a
>> try to the opls force field and I managed to get a stable fatty acid
>> box, reproducing very well the experimental density. Of course the
>> simulation will be much longer but it was just a test system.
>> Ioannis
>>
>>> On Tuesday 22 June 2004 06:27 pm, Derrick Guang Yuh Lee wrote:
>>>> dear gmx-users
>>>>
>>>> right now i am working with a dppc bilayer from dr. tieleman's 
>>>> website
>>>> which requires the ffgmx forcefields, but i have a peptide from my
>>>> boss
>>>> which contains all the hydrogens, which is from experimental data so
>>>> she
>>>> wants all those hydrogen's kept, which i can only get to work and
>>>> keep the
>>>> protons there if i use the ffmgx2 forcefields, and obviously there 
>>>> is
>>>> overlap between the two forcefields. is it at all possible to 
>>>> combine
>>>> the
>>>> two sets of forcefields without compromising either structures 
>>>> and/or
>>>> calculations that will be done?
>>>
>>> Hi Derrick
>>>
>>> In fact ffgmx is a united atom f.f., so not all atoms are explicitly
>>> defined.
>>> If you really want to work with those H's, use OPLS, which is an all
>>> atom
>>> forcefield.
>>> Do not use ffgmx2 for MD, only if you want to do some energy
>>> minimization on a
>>> structure.
>>>
>>> I'm not the right person to say if an all atom forcefield is better
>>> than a
>>> united one, like gromos96 or gmx f.f.
>>> But, the number of good scientific publications using the later one,
>>> is an
>>> indication that we can't really say that working with all atoms is 
>>> the
>>> best
>>> choice.  Do not forget, that the computational time increases very
>>> fast with
>>> the number of atoms present in the system to be studied, and that
>>> united f.f.
>>> were parametrized in that way. No loss of information, I mean.
>>>  And eventually, you can add up the H's later.  But, as I said, I'm
>>> just an
>>> user. ;)
>>>
>>>
>>> Regards,
>>> Nuno
>>>
>>> --
>>> **********************************************
>>> Nuno Ricardo Santos Loureiro da Silva Ferreira
>>> Grupo de Química Biológica
>>> Departamento Química
>>> Universidade de Coimbra
>>> Portugal
>>> www.biolchem.qui.uc.pt
>>> **********************************************
>>> "Do not worry about your difficulties in maths.
>>>  I can assure you that mine are still greater."
>>>                                    emc2
>>>
>>> _______________________________________________
>>> gmx-users mailing list
>>> gmx-users at gromacs.org
>>> http://www.gromacs.org/mailman/listinfo/gmx-users
>>> Please don't post (un)subscribe requests to the list. Use the
>>> www interface or send it to gmx-users-request at gromacs.org.
>>>
>>
>> _______________________________________________
>> gmx-users mailing list
>> gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>>
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
>