On Tue, Aug 25, 2015 at 7:45 PM, Arjun Sharma <arjunsharma83.gmail.com>
wrote:
> Dear Jason,
>
> I’m trying out a simple organic molecule for the explicit solvent CpHMD
> simulation.
>
> H-O=C-O
> |
> CH3-C-CH3
> |
> CH3
>
The titrating site is a carboxylate, but all of Amber's carboxylates
define 4 titrating hydrogens -- syn- and anti- positions on both oxygens.
If you only have a single hydrogen, it could easily get 'stuck' in a
protonation state by virtue of rotating to the anti-position while
deprotonated. The approach taken for AS4 and GL4 correct that. So I would
suggest using the same approach for this residue (although there is a
possibility that using pH-REMD will resolve the issues as well).
> I used partial charges from OPLS and calculated the reference energy
> for the deprotonated form with sander TI method (IGB 2). I added this
> molecule to the residue list in Cpinutil.py and Residues.py (in cpinutils
> dir) as per AMBER 15 manual.
>
I just checked the source code and it turns out that I forgot to update
the Amber manual properly... All of the changes you detail below need to
be added to
$AMBERHOME/AmberTools/src/parmed/chemistry/amber/titratable_residues.py
instead. The two files should be nearly identical (they are functionally
identical, but indentation and spacing is a little different). Sorry for
the confusion.
----------------------------------------------------------------------------------------------------------
> Cpinutil.py :
>
>
> LINE 257: change(parm, 'RADII', ':PTN.O=,.OZ=', 1.3).execute()
>
This adjusts the carboxylate oxygen radii, making them smaller by the same
amount that the ghost hydrogen atoms make it larger when the effective GB
radii are computed. As such, you should only be doing this if you define
your titratable residue to *have* the 4 hydrogens.
Furthermore, if you do adjust the oxygen radii (which is necessary to get
good results in explicit solvent for the 4-hydrogen carboxylate groups),
then you need to derive the reference energy *with these radii*. Running a
TI calculation will get you close, but then you should run a titration of
the model compound to adjust the reference energy to give you the correct
pKa.
>
>
> -------------------------------------------------------------------------------------------
>
> Residue.py:
>
Make these changes to titratable_residues.py instead.
>
> # PTN
> refene1 = _ReferenceEnergy(igb2=0)
> refene1.solvent_energies()
> refene1.dielc2_energies(igb2=0)
> refene1.dielc2.solvent_energies()
> refene2 = _ReferenceEnergy(igb2=-44.8284)
> refene2.solvent_energies()
> refene2.dielc2_energies()
> refene2.dielc2.solvent_energies()
> refene2.set_pKa(5.03, deprotonated=True)
>
> PTN = TitratableResidue('PTN', ['CB', 'HB2', 'HB3', 'HB4', 'CA', 'CB5',
> 'HB6', 'HB7', 'HB8', 'CB9', 'HB10', 'HB11', 'HB12', 'C', 'O', 'OZ', 'H'],
> pka=5.03)
> PTN.add_state(protcnt=1, refene=refene1, # protonated
> charges=[-0.180, 0.600, 0.600, 0.600, 0.000, -0.180, 0.600,
> 0.600, 0.600, -0.180, 0.600, 0.600, 0.600, 0.520, -0.440, -0.530, 0.450])
> PTN.add_state(protcnt=0, refene=refene2, # deprotonated
> charges=[-0.180, 0.600, 0.600, 0.600, -0.100, -0.180, 0.600,
> 0.600, 0.600, -0.180, 0.600, 0.600, 0.600, 0.700, -0.800, -0.800, 0.000])
> PTN.check()
> ———————————----------------------------------------------------------------
> I also added this residue under amino acid list in residues.py file &
> titratable_residues.py (AmberTools/src/parmed/chemistry/amber).
>
> I tried to generate the cpin file for the organic molecule but I run into
> "residue not titratable" cpin error. I’m not sure what other supplementary
> files I need to modify to make it work. I would really appreciate your help.
>
It really should *only* be the titratable_residues.py in parmed (the
cpinutils/residues.py file is not used for much, and I just deleted it in
the developer's version and updated the manual so this confusion will not
happen in the future).
Thanks for the report, and good luck,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Aug 26 2015 - 07:00:03 PDT
