Dear Ray, Dave, and others,
So I think I understand it now, so this is mostly just to confirm that I got it right.
PBSA (when setting radiopt=0) uses the section "%FLAG RADII" in the prmtop file, right?
The LJ radii are defined in these two sections "%FLAG LENNARD_JONES_ACOEF" and "%FLAG LENNARD_JONES_BCOEF", but these are not used by PBSA.
For DELPHI or APBS calculations (I know, why not use PBSA? But please indulge me) you need a PQR file and there are many ways to get this.
1) Use "ambpdb" in AMBER and you get a pqr file with radii from the "%FLAG RADII" section of the prmtop. So this seems okay and in accordance with the PBSA protocol.
2) You can also go the long and complicated way and create a PDB file of a snapshot from your simulation (using e.g. VMD). Then feed this into pdb2pqr and ask for the AMBER radii. However, this will actually give you the LJ radii from the Amber ff94 in the pqr file.
So given Dave's answer I've pasted into the next paragraph below, it would not be "wrong" to use the LJ radii. But the result of the two approaches, will not return the same answer (although they may not be very different either), even though you would *think* that they do since pdb2pqr claim to return the AMBER radii - just not the ones matching the "%FLAG RADII" section in the prmtop.
> You can of course use any set of radii you like, or that you wish to try out.
> In my experience, the bondi radii (or their modified variants) work fairly
> well. You can use "optimized" radii (such as SWANSON) and see how much
> difference it makes. You can even use the LJ radii (not sure if these are
> sigma or rmin). There is no single "right" answer; some programs try to
> guide the user towards a recommended set; others try to provide maximum
> flexibility.
Thanks for all your helpful answers and discussion,
Jesper
On Apr 29, 2012, at 2:52 PM, Ray Luo wrote:
> Note that the solvation radius is NOT LJ radius. These are totally different concepts. So the solvation radii are in a completely different section of the prmtop file, Dave meant the radii in this section are used when radiopt = 0, not those in the LJ coefficient sections.
>
> Ray
> --
> Ray Luo, Ph.D.
>
> On Apr 29, 2012, at 11:19 AM, Jesper Sørensen <jesorensen.ucsd.edu> wrote:
>
>> Hi Dave,
>>
>> Thanks for the answer, that clears some things up.
>> I am still not clear on which radii PBSA uses, because if you set radiopt=0, does it then use the "RADII" or the radii based on the A and B coefficients?
>>> %FLAG LENNARD_JONES_ACOEF
>>
>>
>> Maybe others know this answer too?
>>
>> Best,
>> Jesper
>>
>> On Apr 28, 2012, at 7:25 PM, case wrote:
>>
>>>>
>>>> What does PBSA do when you supply it with a prmtop file? Use the vdW
>>>> radii or the "intrinsic electrostatic radii" (unless of course you ask
>>>> it for TYL06 radii).
>>>
>>> Depends on the radiopt parameter; when radiopt=0, it uses whatever radii are in
>>> the prmtop file.
>>
>> _______________________________________________
>> 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Apr 30 2012 - 00:00:06 PDT
