AMBER Archive (2009)

Subject: Re: [AMBER] RED: Charge derivation of transtion state analogue

From: FyD (
Date: Mon Jul 06 2009 - 01:29:11 CDT

Dear Manoj Singh,

> Thanks for your reply!
> My transition state bound structure was not equliberated and I am running
> further equliberation which will take 2-3 more days, and after that I will
> do the charge fitting. However, here are the things which I am planning to
> do and the confusion I have.
> 1. For the amino acids taking part in reaction, I want to fit the charges of
> sidechain atoms and not the backbone atoms. I am going to extract the
> structure of the species taking part in the reaction (amino acids, water,
> and ligand), will cap the ends of amino acids followed by the procedure you
> have suggested using INTRA-MCC for backbone atoms. But I don't understand
> the procedure and purpose of using INTRA-MCC with capping groups.

You can use the dipeptide ACE-AA-NME (AA = amino acid) & you use
INTRA-MCC for the capping groups set to zero. See for instance:

The idea is to be able to make your complex compatible with the
existing FF libraries i.e. the Amber force field topology database
(FFTopDB), as well as to "insert" your complex within your protein - I
guess your complex belong to a protein, right? Thus, inserting your
complex using a central AA fragment will be quite simple.
When you use an INTRA-MCC set to zero for a chemical group associated
with the R flag "Remove", you force the total charge value for this
group to take a value of zero and with the R flag you remove the
corresponding atoms from the force field library. Thus, you get a FF
library with a total charge value = integer (compatible with the
existing Amber FFTopDB) and a central molecular fragment ready to be
inserted in your protein (using a correct LEaP script where you define
the head/tail of your fragment).

> 2. The way to constrain the distance I know is using 'F' keyword in
> Z-matrix Gaussian input file, however, Ante-RED generates Gaussian input
> file in XYZ format. How can I constrain the distance in XYZ input file
> for Gaussian?

- When I want to constrain a dihedral in Gaussian, I use the
"ModRedundant" mode: Opt=(Tight,ModRedundant)
- after the Cartesian coordinate block jump a line.
- then provide the dihedral constraints between 4 atom numbers "1 2 3
4" involved in the constraint such as:
  D 1 2 3 4 180.00 F
D means dihedral, F means fix, 180.00 the value of the constraint.
- then jump 1 (or 2) lines (I forgot) to close the input.

Check in the Gaussian manual but I guess it should be possible to
constraint a bond using something like:
  B 1 2 1.4 F

regards, Francois

> On Sun, Jul 5, 2009 at 5:39 AM, FyD <> wrote:
>> Dear Manoj Singh,
>> I am trying to study binding properties of one substrate and
>>> its transition state analogue in Amber10.
>>> I derived charges of the substrate in RED-III, using its single
>>> orientation
>>> and single conformation.
>>> For transition state analogue I have change the bond length of partial
>>> breaking bonds and constraint the partial formed bond distance using
>>> distance based restrain, consistence to the previous quantum study for the
>>> reaction mechanism (method is similar to Angw. Chem. Int. Ed. Volume 45
>>> Issue 4 <>,
>>> Pages 653 - 657,
>>> used with other system). I ran the molecular dynamics of the this system
>>> to equilibrate the structure according to the new bond lengths at reaction
>>> center. Now, I want to fit the charges of substrate, water molecule and
>>> the
>>> amino acids taking part in reaction (there are two amino acid and a water
>>> molecule along with substrate forms the reaction center). I want to retain
>>> the bond lengths of partial bands during the charge fitting process.
>>> I will be very thankful if some one can guide me through this charge
>>> fitting process in RED-III.
>> You need to:
>> -1 define the atoms to be involved in the charge derivation process in a
>> PDB file (input of Ante_R.E.D.); let's called it "your complex".
>> -2 execute Ante_R.E.D. to generate the QM input for the geometry
>> optimization step & the P2N file for R.E.D.
>> -3 using the QM input generated by Ante_R.E.D., run the QM geometry
>> optimization step in a standalone approach (i.e. without R.E.D.).
>> -4 You check the P2N files: in particular, are the connectivities defined
>> as you want they reflect the molecular topology of your complex.
>> -5 You rename the file generate in -3 into "Mol_red1.log" + the file
>> generate in -4 into "Mol_red1.p2n". You ran R.E.D.-III.x (or R.E.D. Server)
>> using $OPT_Calc = "Off" & $MEP_Calc = "On"; i.e. using the "mode 2".
>> See
>> * If you want to fix a bond, you need to do it during the geometry
>> optimization step using specific keyword/constraint in the QM program.
>> * If you want to fix some charge values in the fitting step, you need to
>> add the INTRA-MCC keyword in the corresponding P2N file.
>> * You might represent your amino-acids by their dipeptide versions
>> (ACE-AA-NME), to remove the 2 capping groups using the INTRA-MCC keyword and
>> setting constraints to zero for these chemical groups.
>> In fact, you have to apply exactly the same strategy for this complex than
>> that
>> you applied to generate the charges for your transition state structure:
>> there is no fundamental difference. Pay also attention to charge
>> equivalencing for chemically equivalent atoms (or atoms you are going to
>> consider equivalent) in such a complex case (This is defined in the 1st
>> column of atom names in the P2N file format).
>> Do not hesitate to ask more questions as this example is quite
>> complex/interesting...
>> regards, Francois

AMBER mailing list