Hi Jason,
I know and that does make sense, since the changes only pertain to the protein. But then mbondi2 were not made for ligands either - correct?
So should the ligand radii not come from the gaff parameters, or am I misunderstanding something and gaff does not contain/specify its own set of GB radii but instead use the mbondi2 radii?
Also, is mbondi2 the standard PB radii in leap now?
Best, Jesper
On May 2, 2012, at 1:07 PM, Jason Swails <jason.swails.gmail.com> wrote:
> What Hai was saying is that "set default pbradii mbondi3" is identical to
> "set default pbradii mbondi2" for all atoms except 3 atom types [HH(Arg)
> and OE(Glu), OD(Asp)].
>
> Also, the atom radii are NOT stored in the force field files (e.g.,
> gaff.dat, parm99.dat, etc.). They are stored in the LEaP (and ParmEd)
> source code. If you want to use the mbondi3 radii, you should load mbondi3
> at the beginning -- the ligand will be treated as though the radii set was
> mbondi2.
>
> HTH,
> Jason
>
> On Wed, May 2, 2012 at 11:05 AM, Jesper Sørensen <jesorensen.ucsd.edu>wrote:
>
>> Thank you Hai. But it is likely that the ligand radii are coming from the
>> GAFF parameters.
>>
>> I wonder if the correct procedure would probably be to "set default
>> PBradii mbondi3" before loading in the ligand - then only the protein radii
>> should change to mbondi3, without trying to alter the ligand radii as well,
>> which again mbondi3 was likely not parameterized for?
>>
>> Or will this also change the ligand radii?
>>
>> Best,
>> Jesper
>>
>>
>>>
>>> For the ligand, mbondi3 should be exactly the same as mbondi2 since only
>>> radii of HH(Arg) and OE(Glu), OD(Asp) atom type are changed in mbondi3
>> from
>>> mbondi2.
>>>
>>> Best
>>> NHAI
>>>
>>>
>>> On Apr 30, 2012, at 9:02 AM, Ray Luo, Ph.D. wrote:
>>>
>>>> Jesper,
>>>>
>>>> Just a comment on a specific point in your email ...
>>>>
>>>> If you set radiopt=1 and have ligand (GAFF atoms) in the system, PBSA
>>>> would drop a bomb and quit. I believe this is the designed behavior.
>>>> This is because we don't have a "rule" to define the generic atomic
>>>> types in organic molecules. As you can image, the rule would require
>>>> far more simulation data to summarize than that for the proteins/NAs.
>>>>
>>>> So I don't think the scenario in your email would occur.
>>>>
>>>> If you have ligand (GAFFs), the only radii you can use with PBSA right
>>>> now is those in the prmtop file and those can be read in by "radiopt =
>>>> 0".
>>>>
>>>> It's not easy to mix different radii set for proteins/NAs and ligand
>>>> unless you can somehow revise the radii section in the prmtop file.
>>>>
>>>> Ray
>>>>
>>>> On Mon, Apr 30, 2012 at 12:11 AM, Jesper Sørensen <jesorensen.ucsd.edu>
>> wrote:
>>>> ...
>>>>> I know (Ambertools 12, section 8.2.3) that PBSA if you use
>> (radiopt=1), will stick to the Antechamber/GAFF radii for the ligand (and
>> perhaps even the FF radii for the protein) if there are GAFF parameters
>> present in the prmtop file, instead of swithing to the Tan & Luo radii
>> (TYL06 radii). But perhaps this is only a requirement for the TYL06 radii
>> and not mbondi* radii? Can you clarify this a bit for me?
>>>>>
>>>>> Best,
>>>>> Jesper
>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> AMBER mailing list
>>>> AMBER.ambermd.org
>>>> http://lists.ambermd.org/mailman/listinfo/amber
>>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
>
>
> --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Candidate
> 352-392-4032
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 02 2012 - 16:30:02 PDT
