Jan,
Greg gave the exact reply that I would have given if I didn't have to turn off
the phone before takeoff.
You may imagine a number of cases where a "special" rule might appear to fix
things - while probably breaking a number of others. MMFF94 is one of the best
force-fields for small molecules, and I would refrain from carrying out any
modification. I myself applied such a kludge to UFF because it was failing to
optimise cyclopropanones and cyclobutanones. But in that case the failure was
dramatic - it distorted even the correct starting structure upon minimization.
This is the only reason why I finally decided to apply that kludge. This is not
the case for MMFF94: if s correct structure is provided, the correct optimised
geometry is generated. If the starting geometry is ill, the optimiser may be
caught in a local minimum and fail to generate the geometry we expect. This
happens also with QM methods, and I would be more inclined to consider it as a
case of GIGO than a failure of the method.
Cheers,
p.
> On 16 May 2014, at 21:09, Jan Domanski <jan...@gmail.com> wrote:
>
> Absolutely – it would become a misnomer at best. How about something called
> different than MMFF94, like JDFF (completely contrived name, for arguments
> sake ;)) that's using MMFF94/UFF under the hood and adds the amide constraint
> under the hood?
>
>
>
>
>
>
>> On 16 May 2014 19:41, Greg Landrum <greg.land...@gmail.com> wrote:
>>
>>> On Fri, May 16, 2014 at 2:22 PM, Jan Domanski <jan...@gmail.com> wrote:
>>> Yup, David you're correct I think the code is in but Paolo guessed the
>>> intentions of my question.
>>>
>>> I appreciate the point about amides (and other planar groups) in
>>> unconstrained minimization, that's totally fair. Let me put it on its head:
>>> what I would like is the constrained minimization being the default
>>> behavior. Take my molecule, identify amides or other special/planar groups,
>>> add constraints and minimize. Forget the unconstrained minimization – I'm
>>> lazy and want my amides planar without all the code Paolo and I exchanged
>>> in this thread (conformers.py scrip) ;)
>>>
>>> Is this totally outrageous? I'm happy to contribute bits from the
>>> conformers.py Paolo and I concocted.
>>
>> Paolo was very careful to implement MMFF94 as it was originally defined and
>> published. I believe that this is a good thing.
>> Any changes that affect the default behavior of the force field would mean
>> that it's no longer the same thing at all. MMFF94-like is a different beast
>> entirely.
>>
>> I believe that what you're asking for would mean doing exactly this.
>>
>>
>> -greg
>
> ------------------------------------------------------------------------------
> "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
> Instantly run your Selenium tests across 300+ browser/OS combos.
> Get unparalleled scalability from the best Selenium testing platform available
> Simple to use. Nothing to install. Get started now for free."
> http://p.sf.net/sfu/SauceLabs
> _______________________________________________
> Rdkit-discuss mailing list
> Rdkit-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
