help in FEP calculations : analysis/confirmation of results

From: Bruno Luís Pinto de Oliveira (boliveira_at_itn.pt)
Date: Thu Sep 08 2011 - 13:59:23 CDT

Dear All,

I'm did an alchemical FEP transformation of an inhibitor into another to calculate the free energy difference. The calculated values are in good agreement with the experimental values. It seems that the thermodynamic cycle I have used and presented in a previous post is OK.

However I have little experience with FEP and I´m not sure I did everything correct. That´s why I am asking your help again. Moreover I´m doing these calculations by myself without supervision (in my group everyone works with TI/AMBER). I don't have no one to ask for help.

I have a couple of questions:

Can a calculated value of - 0,886 Kcal/mol and a experimental value of - 0,470 Kcal/mol be considered of the same magnitude???

Fep calculates the standard deviation ??? How can I calculate a standard deviation?

I only calculate the foward transformation. However I found in the forum and tutorials that it is also possible to calculate the backward transformation? I need to calculate the backward transformation?? For what??? To see if the results are convergents?

Is possible to visualize the transformation during the FEP simulation using the VMD as it is possible in Amber/sander?

I have similars experimental and calculated values but I want to be sure that I did everything right and it do not happens by chance.

Sorry for the long email.

Any help will be appreciated.

Thank you in advance.

Bruno
..................................................................................................................................................
Bruno L. Oliveira
Radiopharmaceutical Sciences Group
Chemistry Department , ITN
http://www.itn.pt/sec/qui/qir/uk_qir_index.htm
http://www.boliveira.com
________________________________
De: Chirag Vora []
Enviado: segunda-feira, 5 de Setembro de 2011 22:36
Para: Bruno Luís Pinto de Oliveira
Cc: namd-l_at_ks.uiuc.edu
Assunto: Re: namd-l: help in FEP calculations_binding free energy difference between 2 inhibitors

Dear Bruno,
The way you presented "thermodynamic cycle" appears acceptable.

In your case PERTURBATION is disappearance of Re(CO)3. Such method is adopted to predict absolute solvation energy.
See this JCTC article http://pubs.acs.org/doi/abs/10.1021/ct900587b

To have better correlation between experimental and FEP calculated values, you must divide your perturbation into intermediate steps.
In other words state A (lambda=0) to state B (lambda=1) conversion can be expressed through few intermediate steps. A ---> A1 --->A2 --->B and so on.

Please let me know if I got the termodynamic cycle correctly

                                                    DeltaG1
Ligand .... Re(CO)3 : enzyme ---------> Ligand .... Re(CO)3 + enzyme
       | DeltaG4 | DeltaG2
       | |

Ligand : enzyme --------> Ligand + enzyme
                                                   DeltaG3

DDbind = DeltaG1 - Delta3 = DeltaG4 - DeltaG2

The way you presented "thermodynamic cycle" appears acceptable.

I calculate the "transformation" of the Rhenium complex into the ligand in water (DeltaG2 = -31,93 Kcal/mol) and in the protein (DeltaG4 = -31,05 Kcal/mol).

DeltDeltG = DeltaG4 - DeltaG2 = - 0,886 Kcal/mol

The obtained results are in good agreement with the experimental results.

Ligand X have a Ki =178 uM = -5,29 Kcal/mol

Ligand X .... Re(CO)3 ki = 84 uM -5,76 Kcal/mol

DeltDeltG = - 0,470 Kcal/mol (experimental)

My doubt is that in my case there isn´t any transformation as the ligand is exactly the same as the Rhenium complex but minus the Re(CO)3 core.

In my calculations I just did the Re(CO)3 core vanish.... Can I use FEP in this particular case?

 In your case PERTURBATION is disappearance of Re(CO)3. Such method is adopted to predict absolute solvation energy.
See this JCTC article http://pubs.acs.org/doi/abs/10.1021/ct900587b

Did I do something wrong? Can I considerer the results correct?

 To have better correlation between experimental and FEP calculated values, you must divide your perturbation into intermediate steps.
In other words state A (lambda=0) to state B (lambda=1) conversion can be expressed through few intermediate steps. A ---> A1 --->A2 --->B and so on.

Hope that answers your question. For any further query feel free to mail me--_000_12145E4A4635F94DBACC483BCAF132E808EC14FEPHOBOSitnpt_--

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:57:43 CST