[gmx-users] Atom types for phosphate group in OPLS-AA force field

[Justin A. Lemkul]

On 4/26/12 12:15 PM, Delmotte, Antoine wrote:
> Thank you very much for your answer. This has been very helpful. I think I will
> follow your advice and try to find similar molecules to find the parameters.
>
> Could I just ask you if you know of a database where I could look for the opls
> parameters which have calculated for other molecules?
>

There is a large database at http://virtualchemistry.org/index.php but I don't 
know if there will necessarily be anything similar to what you're looking for.

-Justin

> Best regards,
>
> Antoine
>
>
> On 26/04/12 16:55, Justin A. Lemkul wrote:
>>
>>
>> On 4/26/12 11:37 AM, Delmotte, Antoine wrote:
>>> Dear Justin,
>>>
>>> Many thanks for your response. I managed to reduce the number of errors by
>>> trying other atom types, by looking at what seemed most likely to be right from
>>> the ffnonbonded.itp and ffbonded.itp files.
>>>
>>> Unfortunately, I did not manage to find a combination that both made some sense,
>>> and allowed to remove the errors. So I guess these things need to be
>>> parametrized, as you said.
>>>
>>> Could you possibly let me know if there is a relatively easy and straightforward
>>> way to find these parameters? (I guess the answer is probably "no", but it's
>>> always worth asking, I guess...)
>>>
>>
>> You can make a first effort by assigning parameters based on existing,
>> similar, parameters. They may or may not exist.
>>
>>> I also noticed that there is a option "-zero" in grompp which "Sets parameters
>>> for bonded interactions without defaults to zero instead of generating an
>>> error".
>>>
>>> What does it mean exactly? Is it just taking the angles and distances inferred
>>> from the structure as the ones at equilibrium? Is it safe to use it?
>>>
>>
>> What it means is "ignore the problem and simply assign the value of zero for
>> equilibrium lengths, force constants, etc."
>>
>>> I am just a beginner with MD, and I just want to do a quick run to test
>>> something. I have no need for super high accuracy, and the run I want to try
>>> will be very small (maybe 1 or 2 ns, or something like that), and for a very
>>> large system (~80 000 atoms). Would that be unreasonable to use that "-zero"
>>> option (it is tempting, I must say...)?
>>>
>>
>> I say that using -zero is a bad idea. It ignores the problem and whatever
>> results you obtain would be questionable, at best. Parameterization of novel
>> species is an advanced topic. Most people will spend several months developing
>> good parameters for a new residue or molecule.
>>
>> -Justin
>>
>

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

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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