Re: [AMBER] how is pH treated in Amber - other than constant pH simulations

From: Jason Swails <>
Date: Fri, 16 Apr 2010 14:25:49 -0400

You can also run constant pH MD to obtain pKas, and based on your
results set your protonation state for explicit water MD, etc. pKas
are inherently ensemble-averaged properties, so no *single* structure
can give you a reliable pKa for a buried ionizable residue.

Note that constant protonation is NOT constant pH, as Carlos pointed
out. It can be a good approximation if one protonation state is
vastly dominant over the other, but there will always exist members of
the ensemble with a different protonation state. For that reason, it
may be useful to run constant pH MD to see if the constant-protonation
state choice you were going to make is valid.

All the best,

On Fri, Apr 16, 2010 at 12:53 PM, Dean Cuebas
<> wrote:
> Dear Andrew,
> I have found the best approach to protonating my proteins is to use the pKa
> predictions of the H++ server in conjunction with the superior H-bonding
> prediction of reduce at the Molprobity server.
> 1) Use Molprobity to flip Asn, Gln, His residues as needed, since MOST
> proteins from XRC need fixing in this regard.
> 2) Save the flipped but NOT protonated pdb to use as input to the H++ server
> for pKa predictions.
> 3) Continue with the Molprobity server to maximize H-bonding possibilities
> and protonate the protein.
> 4) Using visual inspection of the molprobity output for h-bond and clashes,
> and the pKa predictions of H++ you can come to some reasonably confident
> expectations, especially with regards to the correct state of histidine,
> picking the correct HID or HIE neutral annular tautomer, or the protonated
> HIP.
> Realize that constant pH cannot be used with explicit water MD, so if you
> use a water box, you should do the above.
> Hope this helps!
> Dean
>> From: Carlos Simmerling <>
>> Reply-To: AMBER Mailing List <>
>> Date: Fri, 16 Apr 2010 11:10:42 -0500
>> To: AMBER Mailing List <>
>> Subject: Re: [AMBER] how is pH treated in Amber - other than constant pH
>> simulations
>> simulations in amber are either constant pH or constant protonation.
>> default protonations are reasonable for the isolated amino acids, but
>> pka shifts may occur in proteins. it is possible to calculate pkas for
>> the initial structure, assign protonation states, and keep them. this
>> is ok as long as there are not conformation dependent pka changes that
>> cross your pH.
>> On 4/16/10, Andrew Voronkov <> wrote:
>>> Dear Amber users,
>>> when I am looking for questions about pH and Amber I get mostly information
>>> about constant  pH simulations. But what  pH is supposed to be by default,
>>> without constant pH simulations? Just neutral or what it depends from?
>>> As I understand in Amber  pH is mostly set by protonation state of the
>>> molecules, so if not going to constant pH simulation I can approximately
>>> imitate some pH by setting corresponding protonation state distribution of
>>> amino acids. If physiological pH is required for protein simulation what can
>>> be general recommendation here - instant pH or default pH treatment?
>>> Sincerely yours,
>>> Andrew
>>> _______________________________________________
>>> AMBER mailing list
>> --
>> ===================================================================
>> Carlos L. Simmerling, Ph.D.
>> Professor, Department of Chemistry
>> CMM Bldg, Room G80           Phone: (631) 632-1336   Fax: 632-1555
>> Stony Brook University           E-mail:
>> Stony Brook, NY 11794-5115 Web:
>> ===================================================================
>> _______________________________________________
>> AMBER mailing list
> _______________________________________________
> AMBER mailing list

Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
AMBER mailing list
Received on Fri Apr 16 2010 - 11:30:04 PDT
Custom Search