Re: AMBER: Re: TI tutorial

From: Jiten <jiten.postech.ac.kr>
Date: Wed, 29 Jun 2005 12:46:17 +0900

Dear D. Case,

> I am assuming(?) that you mean the binding free energy for the ligand is 1
> kcal/mol greater or smaller, depending on ALA vs. GLY.

Right

> To get relative free energies of binding, you would need to convert
> ALA->GLY
> in the presence of the ligand, and in its absence. This would involve
> changing both the charges and the vdW parameters.

I would rather opt this option. However, I have a little doubt. Could you
tell me what I am doing is right. (I am sorry for many simple questions )

1. For the changing in the charges - I make the DELTA.charge column such
that the charges of the perturbed state is for GLY. (tmp3.tif is the xleap
editor image) Now do the TI calculations in gas and solvent phase. Find the
difference. Do I need to carry out 12 windows in both the cases ?

2. Again for the change is vdW parameters. (Here is what I confused most).

What should be the starting geometry for this calculation : my originally
equilibrated geometry OR from final geometry of charging TI calculation ?

In your TI tutorial the CHARGE and DELTA.charge column are kept zero as the
system was alreay been neutralized. However, in my case the charges of the
system is with that of GLY (as previously ALA --> GLY charge transformation
are done). Therefore, should I keep both the CHARGE and DELTA.charge column
with that of GLY and zero respectively. (I attach the tmp4.tif file for
this setup in xleap)

Having that, I made the following changes H1--> H0, CT --> H0, HC --> DH
(dummy atom) with the following frcmod.A2G file

         ALA to GLY
MASS
DH 0.000

BOND
H0-DH 340.0 1.090 !CT-HC

ANGLE
DH - H0 - DH 35.0 109.50 !HC-CT-HC
CT - H0 - DH 35.0 109.50 !HC-CT-HC

DIHE
- The default values as given by leap

--
--
NONB
DH     1.000    0.00  !is it right ?
Now do the TI calculations for gas and solvent phase and find the 
difference.
The final free energy difference for ALA--GLY is the addition of step 1 and 
step 2. Is it right ?
> Keep in mind that 1kcal/mol is a pretty small number
As it works well using MM-PBSA, I hope it will work also with TI 
calculation.
Sincerely,
Jiten
----- Original Message ----- 
From: "David A. Case" <case.scripps.edu>
To: <amber.scripps.edu>
Sent: Wednesday, June 29, 2005 10:09 AM
Subject: Re: AMBER: Re: TI tutorial
> On Wed, Jun 29, 2005, Jiten wrote:
>>
>>
>> I have two protein systems bound with the same drug molecule. In one
>> position at the active site of the protein ALA --> GLY are mutated which
>> makes a free energy loss for ~ 1 kcal/mole.
>
> I am assuming(?) that you mean the binding free energy for the ligand is 1
> kcal/mol greater or smaller, depending on ALA vs. GLY.  If you mean 
> something
> else, please disgregard the rest of this message.
>
>> I wish to find out this
>> difference using the TI calculations.
>
> To get relative free energies of binding, you would need to convert 
> ALA->GLY
> in the presence of the ligand, and in its absence.  This would involve
> changing both the charges and the vdW parameters.
>
> If you want absolute binding energies for both complexes, you would need 
> to do
> four free energy calculations (ALA -> nothing with lignad, ALA-> nothing
> without ligand, GLY->nothing with ligand, GLY->nothing without ligand.) 
> Each
> of these four simulations would in turn probably be divided into 
> electrostatic
> and van der Waals contributions.
>
> Keep in mind that 1kcal/mol is a pretty small number.
>
>> (This will need me to have frcmod.DH file for the drug molecule ?)
>
> I don't see how you can perfrom any calculations relevant to drug binding
> without having a force field to represent the drug....
>
> ...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
>
> 

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu

Received on Wed Jun 29 2005 - 04:53:01 PDT