AMBER Archive (2009)

Subject: Re: [AMBER] Some queries about pH simulation

From: Jason Swails (
Date: Fri Oct 09 2009 - 13:48:44 CDT


I'll try to answer your questions in bold below where each one is asked.

>Dear community,
> I am trying to simulate my protein in different
>different pHs. I using the procedure given in Amber9 manual (page 172-177).
>I am facing some problem and in dilemma about am I doing the right
>procedure. I am writing the queries and dilemma serial wise.
>1. Page 172 of manual when I type the following command in tleap
>loadAmberParams frcmod.mod_phipsi.1
>it gives error message
>Could not open file frcmod.mod_phipsi.1: not found
This is because the frcmod.mod_phipsi.1 no longer exists. To the best of my
knowledge, this is due to the fact that these phi/psi torsion modifications
are the basis of the SB modification of the ff99 force field. These
parameters are now present in the leaprc.ff99SB file. Therefore, replace
the commands "source leaprc.ff99" and "loadAmberParams frcmod.mod_phipsi.1"
with the single command "source leaprc.ff99SB".

* >2. I am not following what -states flag dose.
The constant pH method operates by choosing from various protonation
statesaccording to a metropolis monte carlo criteria. These are the
states to
which this flag refers. This flag allows you to set the initial protonation
state of each titratable residue in the CPIN file created by
For instance, the amino acid residues aspartate/aspartic acid and
glutamate/glutamic acid are acidic residues that can have various
protonation states. For reasons listed in Mongan's paper, 5 states were
assigned to these residues. State 0 --> deprotonated, and each oxygen can be
protonated in either the 'syn' or 'anti' conformation, giving rise to the
remaining 4 states. has default initial states that it assigns
to each titratable residue, but depending on the initial conformation of the
protein that you're simulating, you may wish to change the default state
that each titratable residue is placed in. Thus, specifying -state
3,1,0,0,1 puts the first residue in state 3, the second in state 1, the
third in state 0, etc. You can also allow the cpin file to be created with
the default states assigned to each residue and change the state in the
"resstate" array defined in that cpin.

* >3. In the -system flag can we use options other than HEWL. If yes then
>are the options.
This is simply a name for your system. If the name happens to be coded into
either or, then it will provide the experimental data
available in the cpin file, but that will only be used by analysis
(post-processing) scripts such as, and will in no way affect your
sander simulation. I, myself, have never specified the flag. Not
specifying the flag simply applies the label "System: Unknown", with no
adverse effects whatsoever.

* >4. After preparation of topology, coordinate, cspin ect. files I have run
>the MD simulation. I have used the same input file mdin given in
>$AMBERHOME/test/cnstph/Run.>cnstph. But while running sander I needed the
>prmtop option while it is not present in Run.cnstph.
This is because each file specified in the command line program call (sander
-i mdin -o mdout -p prmtop -c inpcrd -r restrt -cpin cpin ... etc.) has a
default name applied to it. If it is not specified, that default name is
assumed. For instance, if no -p prmtop_file_name is supplied, then the
prmtop is assumed to be named "prmtop" in the directory that you're
performing the calculation. This is the assumption made in Run.cnstph since
the topology file is called prmtop in that directory. I'm assuming that you
did not name your topology file prmtop, and you thus had to specify a -p

* >5. I have not solvated my protein. I hope it is not needed since constant
>simulation is done in implicit solvent.
It is a good thing you did not solvate your system, since constant pH
simulations will not run in explicit solvent. In addition, the solvent
method used to parameterize the reference compounds (which are crucial to
proper Boltzmann sampling of protonation states), was igb=2, so great care
must be taken if you intend to use an alternative solvent model (for
instance, igb=7) for your simulations. You would either have to adjust your
reference energies, or confirm that they do not need adjustment by verifying
they each have 0.5 fraction protonation at their specified pKa. (These
reference energies are found in statene array in the cpin)

* >Can anybody please reply to my above querries. Thank you in advance.
>Nicholus Bhattacharjee
>PhD Scholar
>Department of Chemistry
>University of Delhi
>Delhi-110007 (INDIA)
>Phone: 9873098743(M)
>AMBER mailing list
> <>

I hope this sufficiently answered your questions, and good luck!

All the best,
-- Jason Swails
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
AMBER mailing list